Search

We present Segment Anything Model 2 (SAM 2), a foundation model towards solving promptable visual segmentation in images and videos. We build a data engine, which improves model and data via user interaction, to collect the largest video segmentation dataset to date. Our model is a simple transformer architecture with streaming memory for real-time video processing. SAM 2 trained on our data provides strong performance across a wide range of tasks. In video segmentation, we observe better accuracy, using 3x fewer interactions than prior approaches. In image segmentation, our model is more accurate and 6x faster than the Segment Anything Model (SAM). We believe that our data, model, and insights will serve as a significant milestone for video segmentation and related perception tasks. We are releasing a version of our model, the dataset and an interactive demo., Comment: Website: https://ai.meta.com/sam2

Label scarcity problem is the main challenge that hinders the wide application of deep learning systems in automatic cardiovascular diseases (CVDs) detection using electrocardiography (ECG). Tuning pre-trained models alleviates this problem by transferring knowledge learned from large datasets to downstream small datasets. However, bottlenecks in computational efficiency and CVDs detection performance limit its clinical applications. It is difficult to improve the detection performance without significantly sacrificing model computational efficiency. Here, we propose a computation-efficient semi-supervised learning paradigm (FastECG) for robust and computation-efficient CVDs detection using ECG. It enables a robust adaptation of pre-trained models on downstream datasets with limited supervision and high computational efficiency. First, a random-deactivation technique is developed to achieve robust and fast low-rank adaptation of pre-trained weights. Subsequently, we propose a one-shot rank allocation module to determine the optimal ranks for the update matrices of the pre-trained weights. Finally, a lightweight semi-supervised learning pipeline is introduced to enhance model performance by leveraging labeled and unlabeled data with high computational efficiency. Extensive experiments on four downstream ECG datasets demonstrate that FastECG not only outperforms the state-of-the-art methods in multi-label CVDs detection but also consumes fewer GPU footprints, training time, and parameter storage space. As such, this paradigm provides an effective solution for achieving high computational efficiency and robust detection performance in the clinical applications of pre-trained models under limited supervision.

Recent advancements in Retrieval-Augmented Generation have significantly enhanced code completion at the repository level. Various RAG-based code completion systems are proposed based on different design choices. For instance, gaining more effectiveness at the cost of repeating the retrieval-generation process multiple times. However, the indiscriminate use of retrieval in current methods reveals issues in both efficiency and effectiveness, as a considerable portion of retrievals are unnecessary and may introduce unhelpful or even harmful suggestions to code language models. To address these challenges, we introduce CARD, a lightweight critique method designed to provide insights into the necessity of retrievals and select the optimal answer from multiple predictions. CARD can seamlessly integrate into any RAG-based code completion system. Our evaluation shows that CARD saves 21% to 46% times of retrieval for Line completion, 14% to 40% times of retrieval for API completion, and 6% to 46.5% times of retrieval for function completion respectively, while improving the accuracy. CARD reduces latency ranging from 16% to 83%. CARD is generalizable to different LMs, retrievers, and programming languages. It is lightweight with training in few seconds and inference in few milliseconds.

Several open-source systems, such as Flower and NVIDIA FLARE, have been developed in recent years while focusing on different aspects of federated learning (FL). Flower is dedicated to implementing a cohesive approach to FL, analytics, and evaluation. Over time, Flower has cultivated extensive strategies and algorithms tailored for FL application development, fostering a vibrant FL community in research and industry. Conversely, FLARE has prioritized the creation of an enterprise-ready, resilient runtime environment explicitly designed for FL applications in production environments. In this paper, we describe our initial integration of both frameworks and show how they can work together to supercharge the FL ecosystem as a whole. Through the seamless integration of Flower and FLARE, applications crafted within the Flower framework can effortlessly operate within the FLARE runtime environment without necessitating any modifications. This initial integration streamlines the process, eliminating complexities and ensuring smooth interoperability between the two platforms, thus enhancing the overall efficiency and accessibility of FL applications., Comment: Added a figure comparing running a Flower application natively or within FLARE

In the ever-evolving landscape of artificial intelligence (AI) and large language models (LLMs), handling and leveraging data effectively has become a critical challenge. Most state-of-the-art machine learning algorithms are data-centric. However, as the lifeblood of model performance, necessary data cannot always be centralized due to various factors such as privacy, regulation, geopolitics, copyright issues, and the sheer effort required to move vast datasets. In this paper, we explore how federated learning enabled by NVIDIA FLARE can address these challenges with easy and scalable integration capabilities, enabling parameter-efficient and full supervised fine-tuning of LLMs for natural language processing and biopharmaceutical applications to enhance their accuracy and robustness.

Electrocardiography (ECG) is a non-invasive tool for predicting cardiovascular diseases (CVDs). Current ECG-based diagnosis systems show promising performance owing to the rapid development of deep learning techniques. However, the label scarcity problem, the co-occurrence of multiple CVDs and the poor performance on unseen datasets greatly hinder the widespread application of deep learning-based models. Addressing them in a unified framework remains a significant challenge. To this end, we propose a multi-label semi-supervised model (ECGMatch) to recognize multiple CVDs simultaneously with limited supervision. In the ECGMatch, an ECGAugment module is developed for weak and strong ECG data augmentation, which generates diverse samples for model training. Subsequently, a hyperparameter-efficient framework with neighbor agreement modeling and knowledge distillation is designed for pseudo-label generation and refinement, which mitigates the label scarcity problem. Finally, a label correlation alignment module is proposed to capture the co-occurrence information of different CVDs within labeled samples and propagate this information to unlabeled samples. Extensive experiments on four datasets and three protocols demonstrate the effectiveness and stability of the proposed model, especially on unseen datasets. As such, this model can pave the way for diagnostic systems that achieve robust performance on multi-label CVDs prediction with limited supervision.

Modern hierarchical vision transformers have added several vision-specific components in the pursuit of supervised classification performance. While these components lead to effective accuracies and attractive FLOP counts, the added complexity actually makes these transformers slower than their vanilla ViT counterparts. In this paper, we argue that this additional bulk is unnecessary. By pretraining with a strong visual pretext task (MAE), we can strip out all the bells-and-whistles from a state-of-the-art multi-stage vision transformer without losing accuracy. In the process, we create Hiera, an extremely simple hierarchical vision transformer that is more accurate than previous models while being significantly faster both at inference and during training. We evaluate Hiera on a variety of tasks for image and video recognition. Our code and models are available at https://github.com/facebookresearch/hiera., Comment: ICML 2023 Oral version. Code+Models: https://github.com/facebookresearch/hiera

Electroencephalography (EEG) is an objective tool for emotion recognition and shows promising performance. However, the label scarcity problem is a main challenge in this field, which limits the wide application of EEG-based emotion recognition. In this paper, we propose a novel semi-supervised learning framework (EEGMatch) to leverage both labeled and unlabeled EEG data. First, an EEG-Mixup based data augmentation method is developed to generate more valid samples for model learning. Second, a semi-supervised two-step pairwise learning method is proposed to bridge prototype-wise and instance-wise pairwise learning, where the prototype-wise pairwise learning measures the global relationship between EEG data and the prototypical representation of each emotion class and the instance-wise pairwise learning captures the local intrinsic relationship among EEG data. Third, a semi-supervised multi-domain adaptation is introduced to align the data representation among multiple domains (labeled source domain, unlabeled source domain, and target domain), where the distribution mismatch is alleviated. Extensive experiments are conducted on two benchmark databases (SEED and SEED-IV) under a cross-subject leave-one-subject-out cross-validation evaluation protocol. The results show the proposed EEGmatch performs better than the state-of-the-art methods under different incomplete label conditions (with 6.89% improvement on SEED and 1.44% improvement on SEED-IV), which demonstrates the effectiveness of the proposed EEGMatch in dealing with the label scarcity problem in emotion recognition using EEG signals. The source code is available at https://github.com/KAZABANA/EEGMatch.

Federated learning (FL) enables building robust and generalizable AI models by leveraging diverse datasets from multiple collaborators without centralizing the data. We created NVIDIA FLARE as an open-source software development kit (SDK) to make it easier for data scientists to use FL in their research and real-world applications. The SDK includes solutions for state-of-the-art FL algorithms and federated machine learning approaches, which facilitate building workflows for distributed learning across enterprises and enable platform developers to create a secure, privacy-preserving offering for multiparty collaboration utilizing homomorphic encryption or differential privacy. The SDK is a lightweight, flexible, and scalable Python package. It allows researchers to apply their data science workflows in any training libraries (PyTorch, TensorFlow, XGBoost, or even NumPy) in real-world FL settings. This paper introduces the key design principles of NVFlare and illustrates some use cases (e.g., COVID analysis) with customizable FL workflows that implement different privacy-preserving algorithms. Code is available at https://github.com/NVIDIA/NVFlare., Comment: Accepted at the International Workshop on Federated Learning, NeurIPS 2022, New Orleans, USA (https://federated-learning.org/fl-neurips-2022); Revised version v2: added Key Components list, system metrics for homomorphic encryption experiment; Extended v3 for journal submission

Single-view RGB-D human reconstruction with implicit functions is often formulated as per-point classification. Specifically, a set of 3D locations within the view-frustum of the camera are first projected independently onto the image and a corresponding feature is subsequently extracted for each 3D location. The feature of each 3D location is then used to classify independently whether the corresponding 3D point is inside or outside the observed object. This procedure leads to sub-optimal results because correlations between predictions for neighboring locations are only taken into account implicitly via the extracted features. For more accurate results we propose the occupancy planes (OPlanes) representation, which enables to formulate single-view RGB-D human reconstruction as occupancy prediction on planes which slice through the camera's view frustum. Such a representation provides more flexibility than voxel grids and enables to better leverage correlations than per-point classification. On the challenging S3D data we observe a simple classifier based on the OPlanes representation to yield compelling results, especially in difficult situations with partial occlusions due to other objects and partial visibility, which haven't been addressed by prior work., Comment: AAAI2023; Code: https://github.com/Xiaoming-Zhao/oplanes

Optimization within a layer of a deep-net has emerged as a new direction for deep-net layer design. However, there are two main challenges when applying these layers to computer vision tasks: (a) which optimization problem within a layer is useful?; (b) how to ensure that computation within a layer remains efficient? To study question (a), in this work, we propose total variation (TV) minimization as a layer for computer vision. Motivated by the success of total variation in image processing, we hypothesize that TV as a layer provides useful inductive bias for deep-nets too. We study this hypothesis on five computer vision tasks: image classification, weakly supervised object localization, edge-preserving smoothing, edge detection, and image denoising, improving over existing baselines. To achieve these results we had to address question (b): we developed a GPU-based projected-Newton method which is $37\times$ faster than existing solutions., Comment: CVPR 2022

Designing equivariance as an inductive bias into deep-nets has been a prominent approach to build effective models, e.g., a convolutional neural network incorporates translation equivariance. However, incorporating these inductive biases requires knowledge about the equivariance properties of the data, which may not be available, e.g., when encountering a new domain. To address this, we study how to discover interpretable equivariances from data. Specifically, we formulate this discovery process as an optimization problem over a model's parameter-sharing schemes. We propose to use the partition distance to empirically quantify the accuracy of the recovered equivariance. Also, we theoretically analyze the method for Gaussian data and provide a bound on the mean squared gap between the studied discovery scheme and the oracle scheme. Empirically, we show that the approach recovers known equivariances, such as permutations and shifts, on sum of numbers and spatially-invariant data., Comment: AISTATS 2022

In this paper, we investigate statistics on alternating words under correspondence between ``possible reflection paths within several layers of glass'' and ``alternating words''. For $v=(v_1,v_2,\cdots,v_n)\in\mathbb{Z}^{n}$, we say $P$ is a path within $n$ glass plates corresponding to $v$, if $P$ has exactly $v_i$ reflections occurring at the $i^{\rm{th}}$ plate for all $i\in\{1,2,\cdots,n\}$. We give a recursion for the number of paths corresponding to $v$ satisfying $v \in \mathbb{Z}^n$ and $\sum_{i\geq 1} v_i=m$. Also, we establish recursions for statistics around the number of paths corresponding to a given vector $v\in\mathbb{Z}^n$ and a closed form for $n=3$. Finally, we give a equivalent condition for the existence of path corresponding to a given vector $v$., Comment: 13 pages, 17 figures

Extracting detailed 3D information of objects from video data is an important goal for holistic scene understanding. While recent methods have shown impressive results when reconstructing meshes of objects from a single image, results often remain ambiguous as part of the object is unobserved. Moreover, existing image-based datasets for mesh reconstruction don't permit to study models which integrate temporal information. To alleviate both concerns we present SAIL-VOS 3D: a synthetic video dataset with frame-by-frame mesh annotations which extends SAIL-VOS. We also develop first baselines for reconstruction of 3D meshes from video data via temporal models. We demonstrate efficacy of the proposed baseline on SAIL-VOS 3D and Pix3D, showing that temporal information improves reconstruction quality. Resources and additional information are available at http://sailvos.web.illinois.edu., Comment: CVPR 2021 Oral

We propose Chirality Nets, a family of deep nets that is equivariant to the "chirality transform," i.e., the transformation to create a chiral pair. Through parameter sharing, odd and even symmetry, we propose and prove variants of standard building blocks of deep nets that satisfy the equivariance property, including fully connected layers, convolutional layers, batch-normalization, and LSTM/GRU cells. The proposed layers lead to a more data efficient representation and a reduction in computation by exploiting symmetry. We evaluate chirality nets on the task of human pose regression, which naturally exploits the left/right mirroring of the human body. We study three pose regression tasks: 3D pose estimation from video, 2D pose forecasting, and skeleton based activity recognition. Our approach achieves/matches state-of-the-art results, with more significant gains on small datasets and limited-data settings., Comment: Accepted to NeurIPS2019

This study explores extended feature space that is indicative of blood pressure (BP) changes for better estimation of continuous BP in an unobtrusive way. A total of 222 features were extracted from noninvasively acquired electrocardiogram (ECG) and photoplethysmogram (PPG) signals with the subject undergoing coronary angiography and/or percutaneous coronary intervention, during which intra-arterial BP was recorded simultaneously with the subject at rest and while administering drugs to induce BP variations. The association between the extracted features and the BP components, i.e. systolic BP (SBP), diastolic BP (DBP), mean BP (MBP), and pulse pressure (PP) were analyzed and evaluated in terms of correlation coefficient, cross sample entropy, and mutual information, respectively. Results show that the most relevant indicator for both SBP and MBP is the pulse full width at half maximum, and for DBP and PP, the amplitude between the peak of the first derivative of PPG (dPPG) to the valley of the second derivative of PPG (sdPPG) and the time interval between the peak of R wave and the sdPPG, respectively. As potential inputs to either the knowledge-guided model or data-driven method for cuffless BP calibration, the proposed expanded features are expected to improve the estimation accuracy of cuffless BP., Comment: 4 pages, 4 figures, 2 tables

Generative adversarial nets (GANs) and variational auto-encoders have significantly improved our distribution modeling capabilities, showing promise for dataset augmentation, image-to-image translation and feature learning. However, to model high-dimensional distributions, sequential training and stacked architectures are common, increasing the number of tunable hyper-parameters as well as the training time. Nonetheless, the sample complexity of the distance metrics remains one of the factors affecting GAN training. We first show that the recently proposed sliced Wasserstein distance has compelling sample complexity properties when compared to the Wasserstein distance. To further improve the sliced Wasserstein distance we then analyze its `projection complexity' and develop the max-sliced Wasserstein distance which enjoys compelling sample complexity while reducing projection complexity, albeit necessitating a max estimation. We finally illustrate that the proposed distance trains GANs on high-dimensional images up to a resolution of 256x256 easily., Comment: Accepted to CVPR 2019

Video object segmentation is challenging yet important in a wide variety of applications for video analysis. Recent works formulate video object segmentation as a prediction task using deep nets to achieve appealing state-of-the-art performance. Due to the formulation as a prediction task, most of these methods require fine-tuning during test time, such that the deep nets memorize the appearance of the objects of interest in the given video. However, fine-tuning is time-consuming and computationally expensive, hence the algorithms are far from real time. To address this issue, we develop a novel matching based algorithm for video object segmentation. In contrast to memorization based classification techniques, the proposed approach learns to match extracted features to a provided template without memorizing the appearance of the objects. We validate the effectiveness and the robustness of the proposed method on the challenging DAVIS-16, DAVIS-17, Youtube-Objects and JumpCut datasets. Extensive results show that our method achieves comparable performance without fine-tuning and is much more favorable in terms of computational time., Comment: Accepted to ECCV 2018

Unsupervised video segmentation plays an important role in a wide variety of applications from object identification to compression. However, to date, fast motion, motion blur and occlusions pose significant challenges. To address these challenges for unsupervised video segmentation, we develop a novel saliency estimation technique as well as a novel neighborhood graph, based on optical flow and edge cues. Our approach leads to significantly better initial foreground-background estimates and their robust as well as accurate diffusion across time. We evaluate our proposed algorithm on the challenging DAVIS, SegTrack v2 and FBMS-59 datasets. Despite the usage of only a standard edge detector trained on 200 images, our method achieves state-of-the-art results outperforming deep learning based methods in the unsupervised setting. We even demonstrate competitive results comparable to deep learning based methods in the semi-supervised setting on the DAVIS dataset., Comment: Accepted to ECCV 2018

Instance level video object segmentation is an important technique for video editing and compression. To capture the temporal coherence, in this paper, we develop MaskRNN, a recurrent neural net approach which fuses in each frame the output of two deep nets for each object instance -- a binary segmentation net providing a mask and a localization net providing a bounding box. Due to the recurrent component and the localization component, our method is able to take advantage of long-term temporal structures of the video data as well as rejecting outliers. We validate the proposed algorithm on three challenging benchmark datasets, the DAVIS-2016 dataset, the DAVIS-2017 dataset, and the Segtrack v2 dataset, achieving state-of-the-art performance on all of them., Comment: Accepted to NIPS 2017

Existing methods for arterial blood pressure (BP) estimation directly map the input physiological signals to output BP values without explicitly modeling the underlying temporal dependencies in BP dynamics. As a result, these models suffer from accuracy decay over a long time and thus require frequent calibration. In this work, we address this issue by formulating BP estimation as a sequence prediction problem in which both the input and target are temporal sequences. We propose a novel deep recurrent neural network (RNN) consisting of multilayered Long Short-Term Memory (LSTM) networks, which are incorporated with (1) a bidirectional structure to access larger-scale context information of input sequence, and (2) residual connections to allow gradients in deep RNN to propagate more effectively. The proposed deep RNN model was tested on a static BP dataset, and it achieved root mean square error (RMSE) of 3.90 and 2.66 mmHg for systolic BP (SBP) and diastolic BP (DBP) prediction respectively, surpassing the accuracy of traditional BP prediction models. On a multi-day BP dataset, the deep RNN achieved RMSE of 3.84, 5.25, 5.80 and 5.81 mmHg for the 1st day, 2nd day, 4th day and 6th month after the 1st day SBP prediction, and 1.80, 4.78, 5.0, 5.21 mmHg for corresponding DBP prediction, respectively, which outperforms all previous models with notable improvement. The experimental results suggest that modeling the temporal dependencies in BP dynamics significantly improves the long-term BP prediction accuracy., Comment: To appear in IEEE BHI 2018

In this paper, we study learning generalized driving style representations from automobile GPS trip data. We propose a novel Autoencoder Regularized deep neural Network (ARNet) and a trip encoding framework trip2vec to learn drivers' driving styles directly from GPS records, by combining supervised and unsupervised feature learning in a unified architecture. Experiments on a challenging driver number estimation problem and the driver identification problem show that ARNet can learn a good generalized driving style representation: It significantly outperforms existing methods and alternative architectures by reaching the least estimation error on average (0.68, less than one driver) and the highest identification accuracy (by at least 3% improvement) compared with traditional supervised learning methods.

Characterizing driving styles of human drivers using vehicle sensor data, e.g., GPS, is an interesting research problem and an important real-world requirement from automotive industries. A good representation of driving features can be highly valuable for autonomous driving, auto insurance, and many other application scenarios. However, traditional methods mainly rely on handcrafted features, which limit machine learning algorithms to achieve a better performance. In this paper, we propose a novel deep learning solution to this problem, which could be the first attempt of extending deep learning to driving behavior analysis based on GPS data. The proposed approach can effectively extract high level and interpretable features describing complex driving patterns. It also requires significantly less human experience and work. The power of the learned driving style representations are validated through the driver identification problem using a large real dataset.

Grid mapping is a well established approach for environment perception in robotic and automotive applications. Early work suggests estimating the occupancy state of each grid cell in a robot's environment using a Bayesian filter to recursively combine new measurements with the current posterior state estimate of each grid cell. This filter is often referred to as binary Bayes filter (BBF). A basic assumption of classical occupancy grid maps is a stationary environment. Recent publications describe bottom-up approaches using particles to represent the dynamic state of a grid cell and outline prediction-update recursions in a heuristic manner. This paper defines the state of multiple grid cells as a random finite set, which allows to model the environment as a stochastic, dynamic system with multiple obstacles, observed by a stochastic measurement system. It motivates an original filter called the probability hypothesis density / multi-instance Bernoulli (PHD/MIB) filter in a top-down manner. The paper presents a real-time application serving as a fusion layer for laser and radar sensor data and describes in detail a highly efficient parallel particle filter implementation. A quantitative evaluation shows that parameters of the stochastic process model affect the filter results as theoretically expected and that appropriate process and observation models provide consistent state estimation results.

With the aim to improve the performance of feature matching, we present an unsupervised approach to fuse various local descriptors in the space of homographies. Inspired by the observation that the homographies of correct feature correspondences vary smoothly along the spatial domain, our approach stands on the unsupervised nature of feature matching, and can select a good descriptor for matching each feature point. Specifically, the homography space serves as the common domain, in which a correspondence obtained by any descriptor is considered as a point, for integrating various heterogeneous descriptors. Both geometric coherence and spatial continuity among correspondences are considered via computing their geodesic distances in the space. In this way, mutual verification across different descriptors is allowed, and correct correspondences will be highlighted with a high degree of consistency (i.e., short geodesic distances here). It follows that one-class SVM can be applied to identifying these correct correspondences, and boosts the performance of feature matching. The proposed approach is comprehensively compared with the state-of-the-art approaches, and evaluated on four benchmarks of image matching. The promising results manifest its effectiveness.

Using a large-N approach, we demonstrate that the differential conductance and quasi-particle interference pattern measured in recent scanning tunneling spectroscopy experiments (A.R. Schmidt et al. Nature 465, 570 (2010); P. Aynajian et al., PNAS 107, 10383 (2010)) in URu2Si2 are consistent with the emergence of a coherent Kondo lattice below its hidden order transition (HOT). Its formation is driven by a significant increase in the quasi-particle lifetime, which could arise from the emergence of a yet unknown order parameter at the HOT., Comment: 5 pages, 3 figures