Search

In this paper, we investigate the feasibility of leveraging large language models (LLMs) for integrating general knowledge and incorporating pseudo-events as priors for temporal content distribution in video moment retrieval (VMR) models. The motivation behind this study arises from the limitations of using LLMs as decoders for generating discrete textual descriptions, which hinders their direct application to continuous outputs like salience scores and inter-frame embeddings that capture inter-frame relations. To overcome these limitations, we propose utilizing LLM encoders instead of decoders. Through a feasibility study, we demonstrate that LLM encoders effectively refine inter-concept relations in multimodal embeddings, even without being trained on textual embeddings. We also show that the refinement capability of LLM encoders can be transferred to other embeddings, such as BLIP and T5, as long as these embeddings exhibit similar inter-concept similarity patterns to CLIP embeddings. We present a general framework for integrating LLM encoders into existing VMR architectures, specifically within the fusion module. Through experimental validation, we demonstrate the effectiveness of our proposed methods by achieving state-of-the-art performance in VMR. The source code can be accessed at https://github.com/fletcherjiang/LLMEPET., Comment: Accepted to ACM Multimedia 2024

Layout generation is the keystone in achieving automated graphic design, requiring arranging the position and size of various multi-modal design elements in a visually pleasing and constraint-following manner. Previous approaches are either inefficient for large-scale applications or lack flexibility for varying design requirements. Our research introduces a unified framework for automated graphic layout generation, leveraging the multi-modal large language model (MLLM) to accommodate diverse design tasks. In contrast, our data-driven method employs structured text (JSON format) and visual instruction tuning to generate layouts under specific visual and textual constraints, including user-defined natural language specifications. We conducted extensive experiments and achieved state-of-the-art (SOTA) performance on public multi-modal layout generation benchmarks, demonstrating the effectiveness of our method. Moreover, recognizing existing datasets' limitations in capturing the complexity of real-world graphic designs, we propose two new datasets for much more challenging tasks (user-constrained generation and complicated poster), further validating our model's utility in real-life settings. Marking by its superior accessibility and adaptability, this approach further automates large-scale graphic design tasks. The code and datasets will be publicly available on https://github.com/posterllava/PosterLLaVA., Comment: 10 pages; typos corrected, appendix added

Efficient visual perception using mobile systems is crucial, particularly in unknown environments such as search and rescue operations, where swift and comprehensive perception of objects of interest is essential. In such real-world applications, objects of interest are often situated in complex environments, making the selection of the 'Next Best' view based solely on maximizing visibility gain suboptimal. Semantics, providing a higher-level interpretation of perception, should significantly contribute to the selection of the next viewpoint for various perception tasks. In this study, we formulate a novel information gain that integrates both visibility gain and semantic gain in a unified form to select the semantic-aware Next-Best-View. Additionally, we design an adaptive strategy with termination criterion to support a two-stage search-and-acquisition manoeuvre on multiple objects of interest aided by a multi-degree-of-freedoms (Multi-DoFs) mobile system. Several semantically relevant reconstruction metrics, including perspective directivity and region of interest (ROI)-to-full reconstruction volume ratio, are introduced to evaluate the performance of the proposed approach. Simulation experiments demonstrate the advantages of the proposed approach over existing methods, achieving improvements of up to 27.13% for the ROI-to-full reconstruction volume ratio and a 0.88234 average perspective directivity. Furthermore, the planned motion trajectory exhibits better perceiving coverage toward the target.

Video temporal grounding (VTG) is a fine-grained video understanding problem that aims to ground relevant clips in untrimmed videos given natural language queries. Most existing VTG models are built upon frame-wise final-layer CLIP features, aided by additional temporal backbones (e.g., SlowFast) with sophisticated temporal reasoning mechanisms. In this work, we claim that CLIP itself already shows great potential for fine-grained spatial-temporal modeling, as each layer offers distinct yet useful information under different granularity levels. Motivated by this, we propose Reversed Recurrent Tuning ($R^2$-Tuning), a parameter- and memory-efficient transfer learning framework for video temporal grounding. Our method learns a lightweight $R^2$ Block containing only 1.5% of the total parameters to perform progressive spatial-temporal modeling. Starting from the last layer of CLIP, $R^2$ Block recurrently aggregates spatial features from earlier layers, then refines temporal correlation conditioning on the given query, resulting in a coarse-to-fine scheme. $R^2$-Tuning achieves state-of-the-art performance across three VTG tasks (i.e., moment retrieval, highlight detection, and video summarization) on six public benchmarks (i.e., QVHighlights, Charades-STA, Ego4D-NLQ, TACoS, YouTube Highlights, and TVSum) even without the additional backbone, demonstrating the significance and effectiveness of the proposed scheme. Our code is available at https://github.com/yeliudev/R2-Tuning., Comment: ECCV 2024 Camera Ready

Autonomous driving progress relies on large-scale annotated datasets. In this work, we explore the potential of generative models to produce vast quantities of freely-labeled data for autonomous driving applications and present SubjectDrive, the first model proven to scale generative data production in a way that could continuously improve autonomous driving applications. We investigate the impact of scaling up the quantity of generative data on the performance of downstream perception models and find that enhancing data diversity plays a crucial role in effectively scaling generative data production. Therefore, we have developed a novel model equipped with a subject control mechanism, which allows the generative model to leverage diverse external data sources for producing varied and useful data. Extensive evaluations confirm SubjectDrive's efficacy in generating scalable autonomous driving training data, marking a significant step toward revolutionizing data production methods in this field., Comment: Project page: https://subjectdrive.github.io/

Diffusion Transformer (DiT) has emerged as the new trend of generative diffusion models on image generation. In view of extremely slow convergence in typical DiT, recent breakthroughs have been driven by mask strategy that significantly improves the training efficiency of DiT with additional intra-image contextual learning. Despite this progress, mask strategy still suffers from two inherent limitations: (a) training-inference discrepancy and (b) fuzzy relations between mask reconstruction & generative diffusion process, resulting in sub-optimal training of DiT. In this work, we address these limitations by novelly unleashing the self-supervised discrimination knowledge to boost DiT training. Technically, we frame our DiT in a teacher-student manner. The teacher-student discriminative pairs are built on the diffusion noises along the same Probability Flow Ordinary Differential Equation (PF-ODE). Instead of applying mask reconstruction loss over both DiT encoder and decoder, we decouple DiT encoder and decoder to separately tackle discriminative and generative objectives. In particular, by encoding discriminative pairs with student and teacher DiT encoders, a new discriminative loss is designed to encourage the inter-image alignment in the self-supervised embedding space. After that, student samples are fed into student DiT decoder to perform the typical generative diffusion task. Extensive experiments are conducted on ImageNet dataset, and our method achieves a competitive balance between training cost and generative capacity., Comment: CVPR 2024

While we enjoy the richness and informativeness of multimodal data, it also introduces interference and redundancy of information. To achieve optimal domain interpretation with limited resources, we propose CSDNet, a lightweight \textbf{C}ross \textbf{S}hallow and \textbf{D}eep Perception \textbf{Net}work designed to integrate two modalities with less coherence, thereby discarding redundant information or even modality. We implement our CSDNet for Salient Object Detection (SOD) task in robotic perception. The proposed method capitalises on spatial information prescreening and implicit coherence navigation across shallow and deep layers of the depth-thermal (D-T) modality, prioritising integration over fusion to maximise the scene interpretation. To further refine the descriptive capabilities of the encoder for the less-known D-T modalities, we also propose SAMAEP to guide an effective feature mapping to the generalised feature space. Our approach is tested on the VDT-2048 dataset, leveraging the D-T modality outperforms those of SOTA methods using RGB-T or RGB-D modalities for the first time, achieves comparable performance with the RGB-D-T triple-modality benchmark method with 5.97 times faster at runtime and demanding 0.0036 times fewer FLOPs. Demonstrates the proposed CSDNet effectively integrates the information from the D-T modality. The code will be released upon acceptance.

The knowledge of channel covariance matrices is crucial to the design of intelligent reflecting surface (IRS) assisted communication. However, channel covariance matrices may change suddenly in practice. This letter focuses on the detection of the above change in IRS-assisted communication. Specifically, we consider the uplink communication system consisting of a single-antenna user (UE), an IRS, and a multi-antenna base station (BS). We first categorize two types of channel covariance matrix changes based on their impact on system design: Type I change, which denotes the change in the BS receive covariance matrix, and Type II change, which denotes the change in the IRS transmit/receive covariance matrix. Secondly, a powerful method is proposed to detect whether a Type I change occurs, a Type II change occurs, or no change occurs. The effectiveness of our proposed scheme is verified by numerical results., Comment: accepted by IEEE Wireless Communications Letters

With the advancement of video sensors in the Internet of Things, Internet of Video Things (IoVT) systems, capable of delivering abundant and diverse information, have been increasingly deployed for various applications. However, the extensive transmission of video data in IoVT poses challenges in terms of delay and power consumption. Intelligent reconfigurable surface (IRS), as an emerging technology, can enhance communication quality and consequently improve system performance by reconfiguring wireless propagation environments. Inspired by this, we propose a multi-IRS aided IoVT system that leverages IRS to enhance communication quality, thereby reducing power consumption while satisfying delay requirements. To fully leverage the benefits of IRS, we jointly optimize power control for IoVT devices and passive beamforming for IRS to minimize long-term total power consumption under delay constraints. To solve this problem, we first utilize Lyapunov optimization to decouple the long-term optimization problem into each time slot. Subsequently, an alternating optimization algorithm employing optimal solution-seeking and fractional programming is proposed to effectively solve the optimization problems at each time slot. Simulation results demonstrate that the proposed algorithm significantly outperforms benchmark algorithms in terms of long-term total power consumption. Moreover, a trade-off between the number of IRS elements and system performance is also proved.

Hand-drawn 2D animation workflow is typically initiated with the creation of sketch keyframes. Subsequent manual inbetweens are crafted for smoothness, which is a labor-intensive process and the prospect of automatic animation sketch interpolation has become highly appealing. Yet, common frame interpolation methods are generally hindered by two key issues: 1) limited texture and colour details in sketches, and 2) exaggerated alterations between two sketch keyframes. To overcome these issues, we propose a novel deep learning method - Sketch-Aware Interpolation Network (SAIN). This approach incorporates multi-level guidance that formulates region-level correspondence, stroke-level correspondence and pixel-level dynamics. A multi-stream U-Transformer is then devised to characterize sketch inbetweening patterns using these multi-level guides through the integration of self / cross-attention mechanisms. Additionally, to facilitate future research on animation sketch inbetweening, we constructed a large-scale dataset - STD-12K, comprising 30 sketch animation series in diverse artistic styles. Comprehensive experiments on this dataset convincingly show that our proposed SAIN surpasses the state-of-the-art interpolation methods.

The increasing prevalence of mobile devices has led to significant advancements in mobile camera systems and improved image quality. Nonetheless, mobile photography still grapples with challenging issues such as scattering and reflective flare. The absence of a comprehensive real image dataset tailored for mobile phones hinders the development of effective flare mitigation techniques. To address this issue, we present a novel raw image dataset specifically designed for mobile camera systems, focusing on flare removal. Capitalizing on the distinct properties of raw images, this dataset serves as a solid foundation for developing advanced flare removal algorithms. It encompasses a wide variety of real-world scenarios captured with diverse mobile devices and camera settings. The dataset comprises over 2,000 high-quality full-resolution raw image pairs for scattering flare and 1,100 for reflective flare, which can be further segmented into up to 30,000 and 2,200 paired patches, respectively, ensuring broad adaptability across various imaging conditions. Experimental results demonstrate that networks trained with synthesized data struggle to cope with complex lighting settings present in this real image dataset. We also show that processing data through a mobile phone's internal ISP compromises image quality while using raw image data presents significant advantages for addressing the flare removal problem. Our dataset is expected to enable an array of new research in flare removal and contribute to substantial improvements in mobile image quality, benefiting mobile photographers and end-users alike.

3D human pose estimation has been researched for decades with promising fruits. 3D human pose lifting is one of the promising research directions toward the task where both estimated pose and ground truth pose data are used for training. Existing pose lifting works mainly focus on improving the performance of estimated pose, but they usually underperform when testing on the ground truth pose data. We observe that the performance of the estimated pose can be easily improved by preparing good quality 2D pose, such as fine-tuning the 2D pose or using advanced 2D pose detectors. As such, we concentrate on improving the 3D human pose lifting via ground truth data for the future improvement of more quality estimated pose data. Towards this goal, a simple yet effective model called Global-local Adaptive Graph Convolutional Network (GLA-GCN) is proposed in this work. Our GLA-GCN globally models the spatiotemporal structure via a graph representation and backtraces local joint features for 3D human pose estimation via individually connected layers. To validate our model design, we conduct extensive experiments on three benchmark datasets: Human3.6M, HumanEva-I, and MPI-INF-3DHP. Experimental results show that our GLA-GCN implemented with ground truth 2D poses significantly outperforms state-of-the-art methods (e.g., up to around 3%, 17%, and 14% error reductions on Human3.6M, HumanEva-I, and MPI-INF-3DHP, respectively). GitHub: https://github.com/bruceyo/GLA-GCN., Comment: 12 pages, Accepted to ICCV 2023, GitHub code: https://github.com/bruceyo/GLA-GCN

Millimeter-wave(mmWave) technology has emerged as a promising enabler for unleashing the full potential of 360-degree virtual reality (VR). However, the explosive growth of VR services, coupled with the reliability issues of mmWave communications, poses enormous challenges in terms of wireless resource and quality-of-service (QoS) provisioning for mmWave-enabled 360-degree VR. In this paper, we propose an innovative 360-degree VR streaming architecture that addresses three under-exploited issues: overlapping field-of-views (FoVs), statistical QoS provisioning (SQP), and loss-tolerant active data discarding. Specifically, an overlapping FoV-based optimal joint unicast and multicast (JUM) task assignment scheme is designed to implement the non-redundant task assignments, thereby conserving wireless resources remarkably. Furthermore, leveraging stochastic network calculus, we develop a comprehensive SQP theoretical framework that encompasses two SQP schemes from delay and rate perspectives. Additionally, a corresponding optimal adaptive joint time-slot allocation and active-discarding (ADAPT-JTAAT) transmission scheme is proposed to minimize resource consumption while guaranteeing diverse statistical QoS requirements under loss-intolerant and loss-tolerant scenarios from delay and rate perspectives, respectively. Extensive simulations demonstrate the effectiveness of the designed overlapping FoV-based JUM optimal task assignment scheme. Comparisons with six baseline schemes validate that the proposed optimal ADAPTJTAAT transmission scheme can achieve superior SQP performance in resource utilization, flexible rate control, and robust queue behaviors., Comment: 31 pages, 8 figures

Fine-tuning visual models has been widely shown promising performance on many downstream visual tasks. With the surprising development of pre-trained visual foundation models, visual tuning jumped out of the standard modus operandi that fine-tunes the whole pre-trained model or just the fully connected layer. Instead, recent advances can achieve superior performance than full-tuning the whole pre-trained parameters by updating far fewer parameters, enabling edge devices and downstream applications to reuse the increasingly large foundation models deployed on the cloud. With the aim of helping researchers get the full picture and future directions of visual tuning, this survey characterizes a large and thoughtful selection of recent works, providing a systematic and comprehensive overview of existing work and models. Specifically, it provides a detailed background of visual tuning and categorizes recent visual tuning techniques into five groups: prompt tuning, adapter tuning, parameter tuning, and remapping tuning. Meanwhile, it offers some exciting research directions for prospective pre-training and various interactions in visual tuning., Comment: 37 pages. Accepted to ACM CSUR

Predicting the next location is a highly valuable and common need in many location-based services such as destination prediction and route planning. The goal of next location recommendation is to predict the next point-of-interest a user might go to based on the user's historical trajectory. Most existing models learn mobility patterns merely from users' historical check-in sequences while overlooking the significance of user preference modeling. In this work, a novel Point-of-Interest Transformer (POIFormer) with contrastive user preference modeling is developed for end-to-end next location recommendation. This model consists of three major modules: history encoder, query generator, and preference decoder. History encoder is designed to model mobility patterns from historical check-in sequences, while query generator explicitly learns user preferences to generate user-specific intention queries. Finally, preference decoder combines the intention queries and historical information to predict the user's next location. Extensive comparisons with representative schemes and ablation studies on four real-world datasets demonstrate the effectiveness and superiority of the proposed scheme under various settings.

Multi-instance learning (MIL) is an effective paradigm for whole-slide pathological images (WSIs) classification to handle the gigapixel resolution and slide-level label. Prevailing MIL methods primarily focus on improving the feature extractor and aggregator. However, one deficiency of these methods is that the bag contextual prior may trick the model into capturing spurious correlations between bags and labels. This deficiency is a confounder that limits the performance of existing MIL methods. In this paper, we propose a novel scheme, Interventional Bag Multi-Instance Learning (IBMIL), to achieve deconfounded bag-level prediction. Unlike traditional likelihood-based strategies, the proposed scheme is based on the backdoor adjustment to achieve the interventional training, thus is capable of suppressing the bias caused by the bag contextual prior. Note that the principle of IBMIL is orthogonal to existing bag MIL methods. Therefore, IBMIL is able to bring consistent performance boosting to existing schemes, achieving new state-of-the-art performance. Code is available at https://github.com/HHHedo/IBMIL., Comment: Accepted by CVPR 2023; Code at https://github.com/HHHedo/IBMIL

In this paper, fundamentals and performance tradeoffs of the neXt-generation ultra-reliable and low-latency communication (xURLLC) are investigated from the perspective of stochastic network calculus (SNC). An xURLLC-enabled massive MU-MIMO system model has been developed to accommodate xURLLC features. By leveraging and promoting SNC, we provide a quantitative statistical quality of service (QoS) provisioning analysis and derive the closed-form expression of upper-bounded statistical delay violation probability (UB-SDVP). Based on the proposed theoretical framework, we formulate the UB-SDVP minimization problem that is first degenerated into a one-dimensional integer-search problem by deriving the minimum error probability (EP) detector, and then efficiently solved by the integer-form Golden-Section search algorithm. Moreover, two novel concepts, EP-based effective capacity (EP-EC) and EP-based energy efficiency (EP-EE) have been defined to characterize the tail distributions and performance tradeoffs for xURLLC. Subsequently, we formulate the EP-EC and EP-EE maximization problems and prove that the EP-EC maximization problem is equivalent to the UB-SDVP minimization problem, while the EP-EE maximization problem is solved with a low-complexity outer-descent inner-search collaborative algorithm. Extensive simulations demonstrate that the proposed framework in reducing computational complexity compared to reference schemes, and in providing various tradeoffs and optimization performance of xURLLC concerning UB-SDVP, EP, EP-EC, and EP-EE., Comment: 35 pages, 9 figures, Submitted to IEEE TRANSACTIONS ON WIRELESS COMMUNICATIONS

Text-to-motion generation is an emerging and challenging problem, which aims to synthesize motion with the same semantics as the input text. However, due to the lack of diverse labeled training data, most approaches either limit to specific types of text annotations or require online optimizations to cater to the texts during inference at the cost of efficiency and stability. In this paper, we investigate offline open-vocabulary text-to-motion generation in a zero-shot learning manner that neither requires paired training data nor extra online optimization to adapt for unseen texts. Inspired by the prompt learning in NLP, we pretrain a motion generator that learns to reconstruct the full motion from the masked motion. During inference, instead of changing the motion generator, our method reformulates the input text into a masked motion as the prompt for the motion generator to ``reconstruct'' the motion. In constructing the prompt, the unmasked poses of the prompt are synthesized by a text-to-pose generator. To supervise the optimization of the text-to-pose generator, we propose the first text-pose alignment model for measuring the alignment between texts and 3D poses. And to prevent the pose generator from overfitting to limited training texts, we further propose a novel wordless training mechanism that optimizes the text-to-pose generator without any training texts. The comprehensive experimental results show that our method obtains a significant improvement against the baseline methods. The code is available at https://github.com/junfanlin/oohmg.

Parameter efficient transfer learning (PETL) aims at making good use of the representation knowledge in the pre-trained large models by fine-tuning a small number of parameters. Recently, taking inspiration from the natural language processing (NLP) domain, popular PETL techniques such as prompt-tuning and Adapter have also been successfully applied to the vision domain. However, prefix-tuning remains under-explored for vision tasks. In this work, we intend to adapt large vision models (LVMs) to downstream tasks with a good parameter-accuracy trade-off. Towards this goal, we propose a framework with a unified view of PETL called visual-PETL (V-PETL) to investigate the effects of different PETL techniques, data scales of downstream domains, positions of trainable parameters, and other aspects affecting the trade-off. Specifically, we analyze the positional importance of trainable parameters and differences between NLP and vision tasks in terms of data structures and pre-training mechanisms while implementing various PETL techniques, especially for the under-explored prefix-tuning technique. Based on a comprehensive understanding of the differences between NLP and vision data, we propose a new variation of the prefix-tuning module called parallel attention (PATT) for vision downstream tasks. An extensive empirical analysis on vision tasks via different frozen LVMs has been carried and the findings show that the proposed PATT can effectively contribute to other PETL techniques. An effective scheme Swin-BAPAT derived from the proposed V-PETL framework achieves significantly better performance than the state-of-the-art AdaptFormer-Swin with slightly more parameters and outperforms full-tuning with far fewer parameters. Code and data are available at: https://github.com/bruceyo/V-PETL., Comment: under review

The objective of this work is to explore how to effectively and efficiently adapt pre-trained visual foundation models to various downstream tasks of semantic segmentation. Previous methods usually fine-tuned the entire networks for each specific dataset, which will be burdensome to store massive parameters of these networks. A few recent works attempted to insert some extra trainable parameters into the frozen networks to learn visual prompts for parameter-efficient tuning. However, these works showed poor generality as they were designed specifically for Transformers. Moreover, using limited information in these schemes, they exhibited a poor capacity to learn beneficial prompts. To alleviate these issues, we propose a novel Stage-wise Prompt-Matched Framework for generic and effective visual prompt tuning. Specifically, to ensure generality, we divide the pre-trained backbone with frozen parameters into multiple stages and perform prompt learning between different stages, which makes the proposed scheme applicable to various architectures of CNN and Transformer. For effective tuning, a lightweight Semantic-aware Prompt Matcher (SPM) is designed to progressively learn reasonable prompts with a recurrent mechanism, guided by the rich information of interim semantic maps. Working as deep matched filter of representation learning, the proposed SPM can well transform the output of the previous stage into a desirable input for the next stage, thus achieving the better matching/stimulating for the pre-trained knowledge. Extensive experiments on four benchmarks demonstrate that the proposed scheme can achieve a promising trade-off between parameter efficiency and performance effectiveness. Our code and models will be released.

Learned video compression has recently emerged as an essential research topic in developing advanced video compression technologies, where motion compensation is considered one of the most challenging issues. In this paper, we propose a learned video compression framework via heterogeneous deformable compensation strategy (HDCVC) to tackle the problems of unstable compression performance caused by single-size deformable kernels in downsampled feature domain. More specifically, instead of utilizing optical flow warping or single-size-kernel deformable alignment, the proposed algorithm extracts features from the two adjacent frames to estimate content-adaptive heterogeneous deformable (HetDeform) kernel offsets. Then we transform the reference features with the HetDeform convolution to accomplish motion compensation. Moreover, we design a Spatial-Neighborhood-Conditioned Divisive Normalization (SNCDN) to achieve more effective data Gaussianization combined with the Generalized Divisive Normalization. Furthermore, we propose a multi-frame enhanced reconstruction module for exploiting context and temporal information for final quality enhancement. Experimental results indicate that HDCVC achieves superior performance than the recent state-of-the-art learned video compression approaches.

Recent high-performing Human-Object Interaction (HOI) detection techniques have been highly influenced by Transformer-based object detector (i.e., DETR). Nevertheless, most of them directly map parametric interaction queries into a set of HOI predictions through vanilla Transformer in a one-stage manner. This leaves rich inter- or intra-interaction structure under-exploited. In this work, we design a novel Transformer-style HOI detector, i.e., Structure-aware Transformer over Interaction Proposals (STIP), for HOI detection. Such design decomposes the process of HOI set prediction into two subsequent phases, i.e., an interaction proposal generation is first performed, and then followed by transforming the non-parametric interaction proposals into HOI predictions via a structure-aware Transformer. The structure-aware Transformer upgrades vanilla Transformer by encoding additionally the holistically semantic structure among interaction proposals as well as the locally spatial structure of human/object within each interaction proposal, so as to strengthen HOI predictions. Extensive experiments conducted on V-COCO and HICO-DET benchmarks have demonstrated the effectiveness of STIP, and superior results are reported when comparing with the state-of-the-art HOI detectors. Source code is available at \url{https://github.com/zyong812/STIP}., Comment: CVPR 2022; Code is publicly available at: https://github.com/zyong812/STIP

Finding relevant moments and highlights in videos according to natural language queries is a natural and highly valuable common need in the current video content explosion era. Nevertheless, jointly conducting moment retrieval and highlight detection is an emerging research topic, even though its component problems and some related tasks have already been studied for a while. In this paper, we present the first unified framework, named Unified Multi-modal Transformers (UMT), capable of realizing such joint optimization while can also be easily degenerated for solving individual problems. As far as we are aware, this is the first scheme to integrate multi-modal (visual-audio) learning for either joint optimization or the individual moment retrieval task, and tackles moment retrieval as a keypoint detection problem using a novel query generator and query decoder. Extensive comparisons with existing methods and ablation studies on QVHighlights, Charades-STA, YouTube Highlights, and TVSum datasets demonstrate the effectiveness, superiority, and flexibility of the proposed method under various settings. Source code and pre-trained models are available at https://github.com/TencentARC/UMT., Comment: Accepted to Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR 2022)

Translating visual data into natural language is essential for machines to understand the world and interact with humans. In this work, a comprehensive study is conducted on video paragraph captioning, with the goal to generate paragraph-level descriptions for a given video. However, current researches mainly focus on detecting objective facts, ignoring the needs to establish the logical associations between sentences and to discover more accurate emotions related to video contents. Such a problem impairs fluent and abundant expressions of predicted captions, which are far below human language tandards. To solve this problem, we propose to construct a large-scale emotion and logic driven multilingual dataset for this task. This dataset is named EMVPC (standing for "Emotional Video Paragraph Captioning") and contains 53 widely-used emotions in daily life, 376 common scenes corresponding to these emotions, 10,291 high-quality videos and 20,582 elaborated paragraph captions with English and Chinese versions. Relevant emotion categories, scene labels, emotion word labels and logic word labels are also provided in this new dataset. The proposed EMVPC dataset intends to provide full-fledged video paragraph captioning in terms of rich emotions, coherent logic and elaborate expressions, which can also benefit other tasks in vision-language fields. Furthermore, a comprehensive study is conducted through experiments on existing benchmark video paragraph captioning datasets and the proposed EMVPC. The stateof-the-art schemes from different visual captioning tasks are compared in terms of 15 popular metrics, and their detailed objective as well as subjective results are summarized. Finally, remaining problems and future directions of video paragraph captioning are also discussed. The unique perspective of this work is expected to boost further development in video paragraph captioning research.

As a technically challenging topic, visual storytelling aims at generating an imaginary and coherent story with narrative multi-sentences from a group of relevant images. Existing methods often generate direct and rigid descriptions of apparent image-based contents, because they are not capable of exploring implicit information beyond images. Hence, these schemes could not capture consistent dependencies from holistic representation, impairing the generation of reasonable and fluent story. To address these problems, a novel knowledge-enriched attention network with group-wise semantic model is proposed. Three main novel components are designed and supported by substantial experiments to reveal practical advantages. First, a knowledge-enriched attention network is designed to extract implicit concepts from external knowledge system, and these concepts are followed by a cascade cross-modal attention mechanism to characterize imaginative and concrete representations. Second, a group-wise semantic module with second-order pooling is developed to explore the globally consistent guidance. Third, a unified one-stage story generation model with encoder-decoder structure is proposed to simultaneously train and infer the knowledge-enriched attention network, group-wise semantic module and multi-modal story generation decoder in an end-to-end fashion. Substantial experiments on the popular Visual Storytelling dataset with both objective and subjective evaluation metrics demonstrate the superior performance of the proposed scheme as compared with other state-of-the-art methods.

The task of instance segmentation in remote sensing images, aiming at performing per-pixel labeling of objects at instance level, is of great importance for various civil applications. Despite previous successes, most existing instance segmentation methods designed for natural images encounter sharp performance degradations when they are directly applied to top-view remote sensing images. Through careful analysis, we observe that the challenges mainly come from the lack of discriminative object features due to severe scale variations, low contrasts, and clustered distributions. In order to address these problems, a novel context aggregation network (CATNet) is proposed to improve the feature extraction process. The proposed model exploits three lightweight plug-and-play modules, namely dense feature pyramid network (DenseFPN), spatial context pyramid (SCP), and hierarchical region of interest extractor (HRoIE), to aggregate global visual context at feature, spatial, and instance domains, respectively. DenseFPN is a multi-scale feature propagation module that establishes more flexible information flows by adopting inter-level residual connections, cross-level dense connections, and feature re-weighting strategy. Leveraging the attention mechanism, SCP further augments the features by aggregating global spatial context into local regions. For each instance, HRoIE adaptively generates RoI features for different downstream tasks. Extensive evaluations of the proposed scheme on iSAID, DIOR, NWPU VHR-10, and HRSID datasets demonstrate that the proposed approach outperforms state-of-the-arts under similar computational costs. Source code and pre-trained models are available at https://github.com/yeliudev/CATNet., Comment: Accepted to IEEE Transactions on Neural Networks and Learning Systems (TNNLS), 2023

Contrastive learning, which aims at minimizing the distance between positive pairs while maximizing that of negative ones, has been widely and successfully applied in unsupervised feature learning, where the design of positive and negative (pos/neg) pairs is one of its keys. In this paper, we attempt to devise a feature-level data manipulation, differing from data augmentation, to enhance the generic contrastive self-supervised learning. To this end, we first design a visualization scheme for pos/neg score (Pos/neg score indicates cosine similarity of pos/neg pair.) distribution, which enables us to analyze, interpret and understand the learning process. To our knowledge, this is the first attempt of its kind. More importantly, leveraging this tool, we gain some significant observations, which inspire our novel Feature Transformation proposals including the extrapolation of positives. This operation creates harder positives to boost the learning because hard positives enable the model to be more view-invariant. Besides, we propose the interpolation among negatives, which provides diversified negatives and makes the model more discriminative. It is the first attempt to deal with both challenges simultaneously. Experiment results show that our proposed Feature Transformation can improve at least 6.0% accuracy on ImageNet-100 over MoCo baseline, and about 2.0% accuracy on ImageNet-1K over the MoCoV2 baseline. Transferring to the downstream tasks successfully demonstrate our model is less task-bias. Visualization tools and codes https://github.com/DTennant/CL-Visualizing-Feature-Transformation ., Comment: ICCV 2021(Oral), supplementary materials included. Codes and visualization tools: https://github.com/DTennant/CL-Visualizing-Feature-Transformation

Nowadays, Internet of Video Things (IoVT) grows rapidly in terms of quantity and computation demands. In spite of the higher local computation capability on visual processing compared with conventional Internet of Things devices, IoVT devices need to offload partial visual processing tasks to the mobile edge computing (MEC) server wirelessly due to its larger computation demands. However, visual processing task offloading is limited by uplink throughput and computation capability of the MEC server. To break through these limitations, a novel non-orthogonal multiple access (NOMA) assisted IoVT framework with multiple MEC servers is proposed, where NOMA is exploited to improve uplink throughput and MEC servers are co-located with base stations to provide enough computation capability for offloading. In the proposed framework, the association strategy, uplink visual data transmission assisted by NOMA and division of the visual processing tasks as well as computation resource allocation at the MEC servers are jointly optimized to minimize the total delay of all visual processing tasks, while meeting the delay requirements of all IoVT devices. Simulation results demonstrate that significant performance gains can be achieved by proposed joint optimization with NOMA transmission and multi-MEC offloading in the heterogeneous IoVT network., Comment: 8 pages, 5 figures, 15 conference

In multi-cell non-orthogonal multiple access (NOMA) systems, designing an appropriate user grouping strategy is an open problem due to diverse quality of service (QoS) requirements and inter-cell interference. In this paper, we exploit both game theory and graph theory to study QoS-aware user grouping strategies, aiming at minimizing power consumption in downlink multi-cell NOMA systems. Under different QoS requirements, we derive the optimal successive interference cancellation (SIC) decoding order with inter-cell interference, which is different from existing SIC decoding order of increasing channel gains, and obtain the corresponding power allocation strategy. Based on this, the exact potential game model of the user grouping strategies adopted by multiple cells is formulated. We prove that, in this game, the problem for each player to find a grouping strategy can be converted into the problem of searching for specific negative loops in the graph composed of users. Bellman-Ford algorithm is expanded to find these negative loops. Furthermore, we design a greedy based suboptimal strategy to approach the optimal solution with polynomial time. Extensive simulations confirm the effectiveness of grouping users with consideration of QoS and inter-cell interference, and show that the proposed strategies can considerably reduce total power consumption comparing with reference strategies., Comment: 16 pages, 10 figures, 42 conferences, in IEEE Transactions on Wireless Communications, June 2021

The task of classifying mammograms is very challenging because the lesion is usually small in the high resolution image. The current state-of-the-art approaches for medical image classification rely on using the de-facto method for ConvNets - fine-tuning. However, there are fundamental differences between natural images and medical images, which based on existing evidence from the literature, limits the overall performance gain when designed with algorithmic approaches. In this paper, we propose to go beyond fine-tuning by introducing a novel framework called MorphHR, in which we highlight a new transfer learning scheme. The idea behind the proposed framework is to integrate function-preserving transformations, for any continuous non-linear activation neurons, to internally regularise the network for improving mammograms classification. The proposed solution offers two major advantages over the existing techniques. Firstly and unlike fine-tuning, the proposed approach allows for modifying not only the last few layers but also several of the first ones on a deep ConvNet. By doing this, we can design the network front to be suitable for learning domain specific features. Secondly, the proposed scheme is scalable to hardware. Therefore, one can fit high resolution images on standard GPU memory. We show that by using high resolution images, one prevents losing relevant information. We demonstrate, through numerical and visual experiments, that the proposed approach yields to a significant improvement in the classification performance over state-of-the-art techniques, and is indeed on a par with radiology experts. Moreover and for generalisation purposes, we show the effectiveness of the proposed learning scheme on another large dataset, the ChestX-ray14, surpassing current state-of-the-art techniques., Comment: 10 pages, 5 figures

Limited backhaul bandwidth and blockage effects are two main factors limiting the practical deployment of millimeter wave (mmWave) networks. To tackle these issues, we study the feasibility of relaying as well as caching in mmWave networks. A user association and relaying (UAR) criterion dependent on both caching status and maximum biased received power is proposed by considering the spatial correlation caused by the coexistence of base stations (BSs) and relay nodes (RNs). A joint UAR and caching placement problem is then formulated to maximize the backhaul offloading traffic. Using stochastic geometry tools, we decouple the joint UAR and caching placement problem by analyzing the relationship between UAR probabilities and caching placement probabilities. We then optimize the transformed caching placement problem based on polyblock outer approximation by exploiting the monotonic property in the general case and utilizing convex optimization in the noise-limited case. Accordingly, we propose a BS and RN selection algorithm where caching status at BSs and maximum biased received power are jointly considered. Experimental results demonstrate a significant enhancement of backhaul offloading using the proposed algorithms, and show that deploying more RNs and increasing cache size in mmWave networks is a more cost-effective alternative than increasing BS density to achieve similar backhaul offloading performance., Comment: 34 pages, 6 figures

We consider the problem of Human-Object Interaction (HOI) Detection, which aims to locate and recognize HOI instances in the form of in images. Most existing works treat HOIs as individual interaction categories, thus can not handle the problem of long-tail distribution and polysemy of action labels. We argue that multi-level consistencies among objects, actions and interactions are strong cues for generating semantic representations of rare or previously unseen HOIs. Leveraging the compositional and relational peculiarities of HOI labels, we propose ConsNet, a knowledge-aware framework that explicitly encodes the relations among objects, actions and interactions into an undirected graph called consistency graph, and exploits Graph Attention Networks (GATs) to propagate knowledge among HOI categories as well as their constituents. Our model takes visual features of candidate human-object pairs and word embeddings of HOI labels as inputs, maps them into visual-semantic joint embedding space and obtains detection results by measuring their similarities. We extensively evaluate our model on the challenging V-COCO and HICO-DET datasets, and results validate that our approach outperforms state-of-the-arts under both fully-supervised and zero-shot settings. Code is available at https://github.com/yeliudev/ConsNet., Comment: Accepted to Proceedings of the 28th ACM International Conference on Multimedia (MM 2020)

Many semantic events in team sport activities e.g. basketball often involve both group activities and the outcome (score or not). Motion patterns can be an effective means to identify different activities. Global and local motions have their respective emphasis on different activities, which are difficult to capture from the optical flow due to the mixture of global and local motions. Hence it calls for a more effective way to separate the global and local motions. When it comes to the specific case for basketball game analysis, the successful score for each round can be reliably detected by the appearance variation around the basket. Based on the observations, we propose a scheme to fuse global and local motion patterns (MPs) and key visual information (KVI) for semantic event recognition in basketball videos. Firstly, an algorithm is proposed to estimate the global motions from the mixed motions based on the intrinsic property of camera adjustments. And the local motions could be obtained from the mixed and global motions. Secondly, a two-stream 3D CNN framework is utilized for group activity recognition over the separated global and local motion patterns. Thirdly, the basket is detected and its appearance features are extracted through a CNN structure. The features are utilized to predict the success or failure. Finally, the group activity recognition and success/failure prediction results are integrated using the kronecker product for event recognition. Experiments on NCAA dataset demonstrate that the proposed method obtains state-of-the-art performance.

Many important classification problems, such as object classification, speech recognition, and machine translation, have been tackled by the supervised learning paradigm in the past, where training corpora of parallel input-output pairs are required with high cost. To remove the need for the parallel training corpora has practical significance for real-world applications, and it is one of the main goals of unsupervised learning. Recently, encouraging progress in unsupervised learning for solving such classification problems has been made and the nature of the challenges has been clarified. In this article, we review this progress and disseminate a class of promising new methods to facilitate understanding the methods for machine learning researchers. In particular, we emphasize the key information that enables the success of unsupervised learning - the sequential statistics as the distributional prior in the labels. Exploitation of such sequential statistics makes it possible to estimate parameters of classifiers without the need of paired input-output data. In this paper, we first introduce the concept of Caesar Cipher and its decryption, which motivated the construction of the novel loss function for unsupervised learning we use throughout the paper. Then we use a simple but representative binary classification task as an example to derive and describe the unsupervised learning algorithm in a step-by-step, easy-to-understand fashion. We include two cases, one with Bigram language model as the sequential statistics for use in unsupervised parameter estimation, and another with a simpler Unigram language model. For both cases, detailed derivation steps for the learning algorithm are included. Further, a summary table compares computational steps of the two cases in executing the unsupervised learning algorithm for learning binary classifiers.

The learning of Transformation-Equivariant Representations (TERs), which is introduced by Hinton et al. \cite{hinton2011transforming}, has been considered as a principle to reveal visual structures under various transformations. It contains the celebrated Convolutional Neural Networks (CNNs) as a special case that only equivary to the translations. In contrast, we seek to train TERs for a generic class of transformations and train them in an {\em unsupervised} fashion. To this end, we present a novel principled method by Autoencoding Variational Transformations (AVT), compared with the conventional approach to autoencoding data. Formally, given transformed images, the AVT seeks to train the networks by maximizing the mutual information between the transformations and representations. This ensures the resultant TERs of individual images contain the {\em intrinsic} information about their visual structures that would equivary {\em extricably} under various transformations in a generalized {\em nonlinear} case. Technically, we show that the resultant optimization problem can be efficiently solved by maximizing a variational lower-bound of the mutual information. This variational approach introduces a transformation decoder to approximate the intractable posterior of transformations, resulting in an autoencoding architecture with a pair of the representation encoder and the transformation decoder. Experiments demonstrate the proposed AVT model sets a new record for the performances on unsupervised tasks, greatly closing the performance gap to the supervised models., Comment: http://maple-lab.net/projects/AVT.htm. arXiv admin note: text overlap with arXiv:1901.04596

In multi-person videos, especially team sport videos, a semantic event is usually represented as a confrontation between two teams of players, which can be represented as collective motion. In broadcast basketball videos, specific camera motions are used to present specific events. Therefore, a semantic event in broadcast basketball videos is closely related to both the global motion (camera motion) and the collective motion. A semantic event in basketball videos can be generally divided into three stages: pre-event, event occurrence (event-occ), and post-event. In this paper, we propose an ontology-based global and collective motion pattern (On_GCMP) algorithm for basketball event classification. First, a two-stage GCMP based event classification scheme is proposed. The GCMP is extracted using optical flow. The two-stage scheme progressively combines a five-class event classification algorithm on event-occs and a two-class event classification algorithm on pre-events. Both algorithms utilize sequential convolutional neural networks (CNNs) and long short-term memory (LSTM) networks to extract the spatial and temporal features of GCMP for event classification. Second, we utilize post-event segments to predict success/failure using deep features of images in the video frames (RGB_DF_VF) based algorithms. Finally the event classification results and success/failure classification results are integrated to obtain the final results. To evaluate the proposed scheme, we collected a new dataset called NCAA+, which is automatically obtained from the NCAA dataset by extending the fixed length of video clips forward and backward of the corresponding semantic events. The experimental results demonstrate that the proposed scheme achieves the mean average precision of 58.10% on NCAA+. It is higher by 6.50% than state-of-the-art on NCAA.

Non-orthogonal multiple access (NOMA) has shown potential for scalable multicast of video data. However, one key drawback for NOMA-based video multicast is the limited number of layers allowed by the embedded successive interference cancellation algorithm, failing to meet satisfaction of heterogeneous receivers. We propose a novel receiver-driven superposed video multicast (Supcast) scheme by integrating Softcast, an analog-like transmission scheme, into the NOMA-based system to achieve high bandwidth efficiency as well as gradual decoding quality proportional to channel conditions at receivers. Although Softcast allows gradual performance by directly transmitting power-scaled transformation coefficients of frames, it suffers performance degradation due to discarding coefficients under insufficient bandwidth and its power allocation strategy cannot be directly applied in NOMA due to interference. In Supcast, coefficients are grouped into chunks, which are basic units for power allocation and superposition scheduling. By bisecting chunks into base-layer chunks and enhanced-layer chunks, the joint power allocation and chunk scheduling is formulated as a distortion minimization problem. A two-stage power allocation strategy and a near-optimal low-complexity algorithm for chunk scheduling based on the matching theory are proposed. Simulation results have shown the advantage of Supcast against Softcast as well as the reference scheme in NOMA under various practical scenarios., Comment: 9 pages, 6 figures. This paper has already been accepted by IEEE INFOCOM 2019

Unsupervised image translation, which aims in translating two independent sets of images, is challenging in discovering the correct correspondences without paired data. Existing works build upon Generative Adversarial Network (GAN) such that the distribution of the translated images are indistinguishable from the distribution of the target set. However, such set-level constraints cannot learn the instance-level correspondences (e.g. aligned semantic parts in object configuration task). This limitation often results in false positives (e.g. geometric or semantic artifacts), and further leads to mode collapse problem. To address the above issues, we propose a novel framework for instance-level image translation by Deep Attention GAN (DA-GAN). Such a design enables DA-GAN to decompose the task of translating samples from two sets into translating instances in a highly-structured latent space. Specifically, we jointly learn a deep attention encoder, and the instancelevel correspondences could be consequently discovered through attending on the learned instance pairs. Therefore, the constraints could be exploited on both set-level and instance-level. Comparisons against several state-ofthe- arts demonstrate the superiority of our approach, and the broad application capability, e.g, pose morphing, data augmentation, etc., pushes the margin of domain translation problem., Comment: Computer Vision and Pattern Recognition

Exponential growth of social media consumption demands effective user interests-social contents modeling for more personalized recommendation and social media summarization. However, due to the heterogeneous nature of social contents, traditional approaches lack the ability of capturing the hidden semantic correlations across these multi-modal data, which leads to semantic gaps between social content understanding and user interests. To effectively bridge the semantic gaps, we propose a novel deep learning framework for user interests-social contents modeling. We first mine and parse data, i.e. textual content, visual content, social context and social relation, from heterogeneous social media feeds. Then, we design a two-branch network to map the social contents and users into a same latent space. Particularly, the network is trained by a large margin objective that combines a cross-instance distance constraint with a within-instance semantic-preserving constraint in an end-to- end manner. At last, a Deep Semantic-Preserving Embedding (D-Sempre) is learned, and the ranking results can be given by calculating distances between social contents and users. To demonstrate the effectiveness of D-Sempre in user interests-social contents modeling, we construct a Twitter dataset and conduct extensive experiments on it. As a result, D-Sempre effectively integrates the multi-modal data from heterogeneous social media feeds and captures the hidden semantic correlations between users' interests and social contents., Comment: ACM Multimedia

Modeling statistical regularity plays an essential role in ill-posed image processing problems. Recently, deep learning based methods have been presented to implicitly learn statistical representation of pixel distributions in natural images and leverage it as a constraint to facilitate subsequent tasks, such as color constancy and image dehazing. However, the existing CNN architecture is prone to variability and diversity of pixel intensity within and between local regions, which may result in inaccurate statistical representation. To address this problem, this paper presents a novel fully point-wise CNN architecture for modeling statistical regularities in natural images. Specifically, we propose to randomly shuffle the pixels in the origin images and leverage the shuffled image as input to make CNN more concerned with the statistical properties. Moreover, since the pixels in the shuffled image are independent identically distributed, we can replace all the large convolution kernels in CNN with point-wise ($1*1$) convolution kernels while maintaining the representation ability. Experimental results on two applications: color constancy and image dehazing, demonstrate the superiority of our proposed network over the existing architectures, i.e., using 1/10$\sim$1/100 network parameters and computational cost while achieving comparable performance., Comment: 9 pages, 7 figures. To appear in ACM MM 2018

Recently, non-orthogonal multiple access (NOMA) has been proposed to achieve higher spectral efficiency over conventional orthogonal multiple access. Although it has the potential to meet increasing demands of video services, it is still challenging to provide high performance video streaming. In this research, we investigate, for the first time, a multi-user NOMA system design for video transmission. Various NOMA systems have been proposed for data transmission in terms of throughput or reliability. However, the perceived quality, or the quality-of-experience of users, is more critical for video transmission. Based on this observation, we design a quality-driven scalable video transmission framework with cross-layer support for multi-user NOMA. To enable low complexity multi-user NOMA operations, a novel user grouping strategy is proposed. The key features in the proposed framework include the integration of the quality model for encoded video with the physical layer model for NOMA transmission, and the formulation of multi-user NOMA-based video transmission as a quality-driven power allocation problem. As the problem is non-concave, a global optimal algorithm based on the hidden monotonic property and a suboptimal algorithm with polynomial time complexity are developed. Simulation results show that the proposed multi-user NOMA system outperforms existing schemes in various video delivery scenarios., Comment: 9 pages, 6 figures. This paper has already been accepted by IEEE INFOCOM 2018