Search

This paper discusses four facets of the Knowledge Distillation (KD) process for Convolutional Neural Networks (CNNs) and Vision Transformer (ViT) architectures, particularly when executed on edge devices with constrained processing capabilities. First, we conduct a comparative analysis of the KD process between CNNs and ViT architectures, aiming to elucidate the feasibility and efficacy of employing different architectural configurations for the teacher and student, while assessing their performance and efficiency. Second, we explore the impact of varying the size of the student model on accuracy and inference speed, while maintaining a constant KD duration. Third, we examine the effects of employing higher resolution images on the accuracy, memory footprint and computational workload. Last, we examine the performance improvements obtained by fine-tuning the student model after KD to specific downstream tasks. Through empirical evaluations and analyses, this research provides AI practitioners with insights into optimal strategies for maximizing the effectiveness of the KD process on edge devices.

Biometric Verification (BV) systems often exhibit accuracy disparities across different demographic groups, leading to biases in BV applications. Assessing and quantifying these biases is essential for ensuring the fairness of BV systems. However, existing bias evaluation metrics in BV have limitations, such as focusing exclusively on match or non-match error rates, overlooking bias on demographic groups with performance levels falling between the best and worst performance levels, and neglecting the magnitude of the bias present. This paper presents an in-depth analysis of the limitations of current bias evaluation metrics in BV and, through experimental analysis, demonstrates their contextual suitability, merits, and limitations. Additionally, it introduces a novel general-purpose bias evaluation measure for BV, the ``Sum of Group Error Differences (SEDG)''. Our experimental results on controlled synthetic datasets demonstrate the effectiveness of demographic bias quantification when using existing metrics and our own proposed measure. We discuss the applicability of the bias evaluation metrics in a set of simulated demographic bias scenarios and provide scenario-based metric recommendations. Our code is publicly available under \url{https://github.com/alaaobeid/SEDG}.

Social media users often hold several accounts in their effort to multiply the spread of their thoughts, ideas, and viewpoints. In the particular case of objectionable content, users tend to create multiple accounts to bypass the combating measures enforced by social media platforms and thus retain their online identity even if some of their accounts are suspended. User identity linkage aims to reveal social media accounts likely to belong to the same natural person so as to prevent the spread of abusive/illegal activities. To this end, this work proposes a machine learning-based detection model, which uses multiple attributes of users' online activity in order to identify whether two or more virtual identities belong to the same real natural person. The models efficacy is demonstrated on two cases on abusive and terrorism-related Twitter content.

Current progress in the artificial intelligence domain has led to the development of various types of AI-powered dementia assessments, which can be employed to identify patients at the early stage of dementia. It can revolutionize the dementia care settings. It is essential that the medical community be aware of various AI assessments and choose them considering their degrees of validity, efficiency, practicality, reliability, and accuracy concerning the early identification of patients with dementia (PwD). On the other hand, AI developers should be informed about various non-AI assessments as well as recently developed AI assessments. Thus, this paper, which can be readable by both clinicians and AI engineers, fills the gap in the literature in explaining the existing solutions for the recognition of dementia to clinicians, as well as the techniques used and the most widespread dementia datasets to AI engineers. It follows a review of papers on AI and non-AI assessments for dementia to provide valuable information about various dementia assessments for both the AI and medical communities. The discussion and conclusion highlight the most prominent research directions and the maturity of existing solutions., Comment: 49 pages

New advancements for the detection of synthetic images are critical for fighting disinformation, as the capabilities of generative AI models continuously evolve and can lead to hyper-realistic synthetic imagery at unprecedented scale and speed. In this paper, we focus on the challenge of generalizing across different concept classes, e.g., when training a detector on human faces and testing on synthetic animal images - highlighting the ineffectiveness of existing approaches that randomly sample generated images to train their models. By contrast, we propose an approach based on the premise that the robustness of the detector can be enhanced by training it on realistic synthetic images that are selected based on their quality scores according to a probabilistic quality estimation model. We demonstrate the effectiveness of the proposed approach by conducting experiments with generated images from two seminal architectures, StyleGAN2 and Latent Diffusion, and using three different concepts for each, so as to measure the cross-concept generalization ability. Our results show that our quality-based sampling method leads to higher detection performance for nearly all concepts, improving the overall effectiveness of the synthetic image detectors.

The recent wave of AI research has enabled a new brand of synthetic media, called deepfakes. Deepfakes have impressive photorealism, which has generated exciting new use cases but also raised serious threats to our increasingly digital world. To mitigate these threats, researchers have tried to come up with new methods for deepfake detection that are more effective than traditional forensics and heavily rely on deep AI technology. In this paper, following up on encouraging prior work for deepfake detection with attribution and ensemble techniques, we explore and compare multiple designs for ensemble detectors. The goal is to achieve robustness and good generalization ability by leveraging ensembles of models that specialize in different manipulation categories. Our results corroborate that ensembles can achieve higher accuracy than individual models when properly tuned, while the generalization ability relies on access to a large number of training data for a diverse set of known manipulations., Comment: 16 pages, 4 figures

We introduce S$^2$VS, a video similarity learning approach with self-supervision. Self-Supervised Learning (SSL) is typically used to train deep models on a proxy task so as to have strong transferability on target tasks after fine-tuning. Here, in contrast to prior work, SSL is used to perform video similarity learning and address multiple retrieval and detection tasks at once with no use of labeled data. This is achieved by learning via instance-discrimination with task-tailored augmentations and the widely used InfoNCE loss together with an additional loss operating jointly on self-similarity and hard-negative similarity. We benchmark our method on tasks where video relevance is defined with varying granularity, ranging from video copies to videos depicting the same incident or event. We learn a single universal model that achieves state-of-the-art performance on all tasks, surpassing previously proposed methods that use labeled data. The code and pretrained models are publicly available at: https://github.com/gkordo/s2vs

This Signal Processing Grand Challenge (SPGC) targets a difficult automatic prediction problem of societal and medical relevance, namely, the detection of Alzheimer's Dementia (AD). Participants were invited to employ signal processing and machine learning methods to create predictive models based on spontaneous speech data. The Challenge has been designed to assess the extent to which predictive models built based on speech in one language (English) generalise to another language (Greek). To the best of our knowledge no work has investigated acoustic features of the speech signal in multilingual AD detection. Our baseline system used conventional machine learning algorithms with Active Data Representation of acoustic features, achieving accuracy of 73.91% on AD detection, and 4.95 root mean squared error on cognitive score prediction., Comment: ICASSP 2023 SPGC description

In this paper, we address the problem of image splicing localization with a multi-stream network architecture that processes the raw RGB image in parallel with other handcrafted forensic signals. Unlike previous methods that either use only the RGB images or stack several signals in a channel-wise manner, we propose an encoder-decoder architecture that consists of multiple encoder streams. Each stream is fed with either the tampered image or handcrafted signals and processes them separately to capture relevant information from each one independently. Finally, the extracted features from the multiple streams are fused in the bottleneck of the architecture and propagated to the decoder network that generates the output localization map. We experiment with two handcrafted algorithms, i.e., DCT and Splicebuster. Our proposed approach is benchmarked on three public forensics datasets, demonstrating competitive performance against several competing methods and achieving state-of-the-art results, e.g., 0.898 AUC on CASIA., Comment: Accepted to the International Conference on MultiMedia Modeling (MMM 2023)

For fashion outfits to be considered aesthetically pleasing, the garments that constitute them need to be compatible in terms of visual aspects, such as style, category and color. Previous works have defined visual compatibility as a binary classification task with items in a garment being considered as fully compatible or fully incompatible. However, this is not applicable to Outfit Maker applications where users create their own outfits and need to know which specific items may be incompatible with the rest of the outfit. To address this, we propose the Visual InCompatibility TransfORmer (VICTOR) that is optimized for two tasks: 1) overall compatibility as regression and 2) the detection of mismatching items and utilize fashion-specific contrastive language-image pre-training for fine tuning computer vision neural networks on fashion imagery. We build upon the Polyvore outfit benchmark to generate partially mismatching outfits, creating a new dataset termed Polyvore-MISFITs, that is used to train VICTOR. A series of ablation and comparative analyses show that the proposed architecture can compete and even surpass the current state-of-the-art on Polyvore datasets while reducing the instance-wise floating operations by 88%, striking a balance between high performance and efficiency. We release our code at https://github.com/stevejpapad/Visual-InCompatibility-Transformer, Comment: 14 pages, 6 figures

In this paper we introduce InDistill, a model compression approach that combines knowledge distillation and channel pruning in a unified framework for the transfer of the critical information flow paths from a heavyweight teacher to a lightweight student. Such information is typically collapsed in previous methods due to an encoding stage prior to distillation. By contrast, InDistill leverages a pruning operation applied to the teacher's intermediate layers reducing their width to the corresponding student layers' width. In that way, we force architectural alignment enabling the intermediate layers to be directly distilled without the need of an encoding stage. Additionally, a curriculum learning-based training scheme is adopted considering the distillation difficulty of each layer and the critical learning periods in which the information flow paths are created. The proposed method surpasses state-of-the-art performance on three standard benchmarks, i.e. CIFAR-10, CUB-200, and FashionMNIST by 3.08%, 14.27%, and 1% mAP, respectively, as well as on more challenging evaluation settings, i.e. ImageNet and CIFAR-100 by 1.97% and 5.65% mAP, respectively.

Decentralization is emerging as a key feature of the future Internet. However, effective algorithms for search are missing from state-of-the-art decentralized technologies, such as distributed hash tables and blockchain. This is surprising, since decentralized search has been studied extensively in earlier peer-to-peer (P2P) literature. In this work, we adopt a fresh outlook for decentralized search in P2P networks that is inspired by advancements in dense information retrieval and graph signal processing. In particular, we generate latent representations of P2P nodes based on their stored documents and diffuse them to the rest of the network with graph filters, such as personalized PageRank. We then use the diffused representations to guide search queries towards relevant content. Our preliminary approach is successful in locating relevant documents in nearby nodes but the accuracy declines sharply with the number of stored documents, highlighting the need for more sophisticated techniques., Comment: 7 pages, 3 figures, to be published in the Decentralized Internet, Networks, Protocols, and Systems (DINPS 2022) workshop hosted at the 42nd IEEE International Conference on Distributed Computing Systems (ICDCS 2022) - https://research.protocol.ai/sites/dinps/

Enabled by recent improvements in generation methodologies, DeepFakes have become mainstream due to their increasingly better visual quality, the increase in easy-to-use generation tools and the rapid dissemination through social media. This fact poses a severe threat to our societies with the potential to erode social cohesion and influence our democracies. To mitigate the threat, numerous DeepFake detection schemes have been introduced in the literature but very few provide a web service that can be used in the wild. In this paper, we introduce the MeVer DeepFake detection service, a web service detecting deep learning manipulations in images and video. We present the design and implementation of the proposed processing pipeline that involves a model ensemble scheme, and we endow the service with a model card for transparency. Experimental results show that our service performs robustly on the three benchmark datasets while being vulnerable to Adversarial Attacks. Finally, we outline our experience and lessons learned when deploying a research system into production in the hopes that it will be useful to other academic and industry teams., Comment: 10 pages, 6 figures

Estimating the preferences of consumers is of utmost importance for the fashion industry as appropriately leveraging this information can be beneficial in terms of profit. Trend detection in fashion is a challenging task due to the fast pace of change in the fashion industry. Moreover, forecasting the visual popularity of new garment designs is even more demanding due to lack of historical data. To this end, we propose MuQAR, a Multimodal Quasi-AutoRegressive deep learning architecture that combines two modules: (1) a multi-modal multi-layer perceptron processing categorical, visual and textual features of the product and (2) a quasi-autoregressive neural network modelling the "target" time series of the product's attributes along with the "exogenous" time series of all other attributes. We utilize computer vision, image classification and image captioning, for automatically extracting visual features and textual descriptions from the images of new products. Product design in fashion is initially expressed visually and these features represent the products' unique characteristics without interfering with the creative process of its designers by requiring additional inputs (e.g manually written texts). We employ the product's target attributes time series as a proxy of temporal popularity patterns, mitigating the lack of historical data, while exogenous time series help capture trends among interrelated attributes. We perform an extensive ablation analysis on two large scale image fashion datasets, Mallzee and SHIFT15m to assess the adequacy of MuQAR and also use the Amazon Reviews: Home and Kitchen dataset to assess generalisability to other domains. A comparative study on the VISUELLE dataset, shows that MuQAR is capable of competing and surpassing the domain's current state of the art by 4.65% and 4.8% in terms of WAPE and MAE respectively., Comment: 14 pages, 4 figures

In this work, we aim to classify nodes of unstructured peer-to-peer networks with communication uncertainty, such as users of decentralized social networks. Graph Neural Networks (GNNs) are known to improve the accuracy of simple classifiers in centralized settings by leveraging naturally occurring network links, but graph convolutional layers are challenging to implement in decentralized settings when node neighbors are not constantly available. We address this problem by employing decoupled GNNs, where base classifier predictions and errors are diffused through graphs after training. For these, we deploy pre-trained and gossip-trained base classifiers and implement peer-to-peer graph diffusion under communication uncertainty. In particular, we develop an asynchronous decentralized formulation of diffusion that converges to centralized predictions in distribution and linearly with respect to communication rates. We experiment on three real-world graphs with node features and labels and simulate peer-to-peer networks with uniformly random communication frequencies; given a portion of known labels, our decentralized graph diffusion achieves comparable accuracy to centralized GNNs with minimal communication overhead (less than 3% of what gossip training already adds)., Comment: 12 pages, 4 figures, 2 tables, accepted manuscript, IEEE Access

We introduce pygrank, an open source Python package to define, run and evaluate node ranking algorithms. We provide object-oriented and extensively unit-tested algorithm components, such as graph filters, post-processors, measures, benchmarks and online tuning. Computations can be delegated to numpy, tensorflow or pytorch backends and fit in back-propagation pipelines. Classes can be combined to define interoperable complex algorithms. Within the context of this paper we compare the package with related alternatives and demonstrate its flexibility and ease of use with code examples., Comment: 6 pages, 1 figure, 2 tables, 3 code snippets

Graph filters are an emerging paradigm that systematizes information propagation in graphs as transformation of prior node values, called graph signals, to posterior scores. In this work, we study the problem of mitigating disparate impact, i.e. posterior score differences between a protected set of sensitive nodes and the rest, while minimally editing scores to preserve recommendation quality. To this end, we develop a scheme that respects propagation mechanisms by editing graph signal priors according to their posteriors and node sensitivity, where a small number of editing parameters can be tuned to constrain or eliminate disparate impact. We also theoretically explain that coarse prior editing can locally optimize posteriors objectives thanks to graph filter robustness. We experiment on a diverse collection of 12 graphs with varying number of nodes, where our approach performs equally well or better than previous ones in minimizing disparate impact and preserving posterior AUC under fairness constraints., Comment: 40 pages, 4 figures, 9 tables

Computer Vision (CV) has achieved remarkable results, outperforming humans in several tasks. Nonetheless, it may result in significant discrimination if not handled properly as CV systems highly depend on the data they are fed with and can learn and amplify biases within such data. Thus, the problems of understanding and discovering biases are of utmost importance. Yet, there is no comprehensive survey on bias in visual datasets. Hence, this work aims to: i) describe the biases that might manifest in visual datasets; ii) review the literature on methods for bias discovery and quantification in visual datasets; iii) discuss existing attempts to collect bias-aware visual datasets. A key conclusion of our study is that the problem of bias discovery and quantification in visual datasets is still open, and there is room for improvement in terms of both methods and the range of biases that can be addressed. Moreover, there is no such thing as a bias-free dataset, so scientists and practitioners must become aware of the biases in their datasets and make them explicit. To this end, we propose a checklist to spot different types of bias during visual dataset collection.

In this paper, we address the problem of high performance and computationally efficient content-based video retrieval in large-scale datasets. Current methods typically propose either: (i) fine-grained approaches employing spatio-temporal representations and similarity calculations, achieving high performance at a high computational cost or (ii) coarse-grained approaches representing/indexing videos as global vectors, where the spatio-temporal structure is lost, providing low performance but also having low computational cost. In this work, we propose a Knowledge Distillation framework, called Distill-and-Select (DnS), that starting from a well-performing fine-grained Teacher Network learns: a) Student Networks at different retrieval performance and computational efficiency trade-offs and b) a Selector Network that at test time rapidly directs samples to the appropriate student to maintain both high retrieval performance and high computational efficiency. We train several students with different architectures and arrive at different trade-offs of performance and efficiency, i.e., speed and storage requirements, including fine-grained students that store/index videos using binary representations. Importantly, the proposed scheme allows Knowledge Distillation in large, unlabelled datasets -- this leads to good students. We evaluate DnS on five public datasets on three different video retrieval tasks and demonstrate a) that our students achieve state-of-the-art performance in several cases and b) that the DnS framework provides an excellent trade-off between retrieval performance, computational speed, and storage space. In specific configurations, the proposed method achieves similar mAP with the teacher but is 20 times faster and requires 240 times less storage space. The collected dataset and implementation are publicly available: https://github.com/mever-team/distill-and-select.

In this paper, we address the problem of global-scale image geolocation, proposing a mixed classification-retrieval scheme. Unlike other methods that strictly tackle the problem as a classification or retrieval task, we combine the two practices in a unified solution leveraging the advantages of each approach with two different modules. The first leverages the EfficientNet architecture to assign images to a specific geographic cell in a robust way. The second introduces a new residual architecture that is trained with contrastive learning to map input images to an embedding space that minimizes the pairwise geodesic distance of same-location images. For the final location estimation, the two modules are combined with a search-within-cell scheme, where the locations of most similar images from the predicted geographic cell are aggregated based on a spatial clustering scheme. Our approach demonstrates very competitive performance on four public datasets, achieving new state-of-the-art performance in fine granularity scales, i.e., 15.0% at 1km range on Im2GPS3k.

In this work, we present a deep learning-based approach for image tampering localization fusion. This approach is designed to combine the outcomes of multiple image forensics algorithms and provides a fused tampering localization map, which requires no expert knowledge and is easier to interpret by end users. Our fusion framework includes a set of five individual tampering localization methods for splicing localization on JPEG images. The proposed deep learning fusion model is an adapted architecture, initially proposed for the image restoration task, that performs multiple operations in parallel, weighted by an attention mechanism to enable the selection of proper operations depending on the input signals. This weighting process can be very beneficial for cases where the input signal is very diverse, as in our case where the output signals of multiple image forensics algorithms are combined. Evaluation in three publicly available forensics datasets demonstrates that the performance of the proposed approach is competitive, outperforming the individual forensics techniques as well as another recently proposed fusion framework in the majority of cases., Comment: 6 pages, 3 figures, cbmi