To balance the annotation labor and the granularity of supervision, single-frame annotation has been introduced in temporal action localization. It provides a rough temporal location for an action but implicitly overstates the supervision from the annotated-frame during training, leading to the confusion between actions and backgrounds, i.e., action incompleteness and background false positives. To tackle the two challenges, in this work, we present the Snippet Classification model and the Dilation-Erosion module. In the Dilation-Erosion module, we expand the potential action segments with a loose criterion to alleviate the problem of action incompleteness and then remove the background from the potential action segments to alleviate the problem of action incompleteness. Relying on the single-frame annotation and the output of the snippet classification, the Dilation-Erosion module mines pseudo snippet-level ground-truth, hard backgrounds and evident backgrounds, which in turn further trains the Snippet Classification model. It forms a cyclic dependency. Furthermore, we propose a new embedding loss to aggregate the features of action instances with the same label and separate the features of actions from backgrounds. Experiments on THUMOS14 and ActivityNet 1.2 validate the effectiveness of the proposed method. Code has been made publicly available (https://github.com/LingJun123/single-frame-TAL).
The foundation models have recently shown excellent performance on a variety of downstream tasks in computer vision. However, most existing vision foundation models simply focus on image-level pretraining and adpation, which are limited for dynamic and complex video-level understanding tasks. To fill the gap, we present general video foundation models, InternVideo, by taking advantage of both generative and discriminative self-supervised video learning. Specifically, InternVideo efficiently explores masked video modeling and video-language contrastive learning as the pretraining objectives, and selectively coordinates video representations of these two complementary frameworks in a learnable manner to boost various video applications. Without bells and whistles, InternVideo achieves state-of-the-art performance on 39 video datasets from extensive tasks including video action recognition/detection, video-language alignment, and open-world video applications. Especially, our methods can obtain 91.1% and 77.2% top-1 accuracy on the challenging Kinetics-400 and Something-Something V2 benchmarks, respectively. All of these results effectively show the generality of our InternVideo for video understanding. The code will be released at https://github.com/OpenGVLab/InternVideo .
Vision Transformers (ViTs) have achieved overwhelming success, yet they suffer from vulnerable resolution scalability, i.e., the performance drops drastically when presented with input resolutions that are unseen during training. We introduce, ResFormer, a framework that is built upon the seminal idea of multi-resolution training for improved performance on a wide spectrum of, mostly unseen, testing resolutions. In particular, ResFormer operates on replicated images of different resolutions and enforces a scale consistency loss to engage interactive information across different scales. More importantly, to alternate among varying resolutions effectively, especially novel ones in testing, we propose a global-local positional embedding strategy that changes smoothly conditioned on input sizes. We conduct extensive experiments for image classification on ImageNet. The results provide strong quantitative evidence that ResFormer has promising scaling abilities towards a wide range of resolutions. For instance, ResFormer-B-MR achieves a Top-1 accuracy of 75.86% and 81.72% when evaluated on relatively low and high resolutions respectively (i.e., 96 and 640), which are 48% and 7.49% better than DeiT-B. We also demonstrate, moreover, ResFormer is flexible and can be easily extended to semantic segmentation, object detection and video action recognition. Code is available at https://github.com/ruitian12/resformer.
Although action recognition systems can achieve top performance when evaluated on in-distribution test points, they are vulnerable to unanticipated distribution shifts in test data. However, test-time adaptation of video action recognition models against common distribution shifts has so far not been demonstrated. We propose to address this problem with an approach tailored to spatio-temporal models that is capable of adaptation on a single video sample at a step. It consists in a feature distribution alignment technique that aligns online estimates of test set statistics towards the training statistics. We further enforce prediction consistency over temporally augmented views of the same test video sample. Evaluations on three benchmark action recognition datasets show that our proposed technique is architecture-agnostic and able to significantly boost the performance on both, the state of the art convolutional architecture TANet and the Video Swin Transformer. Our proposed method demonstrates a substantial performance gain over existing test-time adaptation approaches in both evaluations of a single distribution shift and the challenging case of random distribution shifts. Code will be available at \url{https://github.com/wlin-at/ViTTA}.
Temporal action localization (TAL) requires long-form reasoning to predict actions of various durations and complex content. Given limited GPU memory, training TAL end to end (i.e., from videos to predictions) on long videos is a significant challenge. Most methods can only train on pre-extracted features without optimizing them for the localization problem, consequently limiting localization performance. In this work, to extend the potential in TAL networks, we propose a novel end-to-end method Re2TAL, which rewires pretrained video backbones for reversible TAL. Re2TAL builds a backbone with reversible modules, where the input can be recovered from the output such that the bulky intermediate activations can be cleared from memory during training. Instead of designing one single type of reversible module, we propose a network rewiring mechanism, to transform any module with a residual connection to a reversible module without changing any parameters. This provides two benefits: (1) a large variety of reversible networks are easily obtained from existing and even future model designs, and (2) the reversible models require much less training effort as they reuse the pre-trained parameters of their original non-reversible versions. Re2TAL, only using the RGB modality, reaches 37.01% average mAP on ActivityNet-v1.3, a new state-of-the-art record, and mAP 64.9% at tIoU=0.5 on THUMOS-14, outperforming all other RGB-only methods.
In video action recognition, shortcut static features can interfere with the learning of motion features, resulting in poor out-of-distribution (OOD) generalization. The video background is clearly a source of static bias, but the video foreground, such as the clothing of the actor, can also provide static bias. In this paper, we empirically verify the existence of foreground static bias by creating test videos with conflicting signals from the static and moving portions of the video. To tackle this issue, we propose a simple yet effective technique, StillMix, to learn robust action representations. Specifically, StillMix identifies bias-inducing video frames using a 2D reference network and mixes them with videos for training, serving as effective bias suppression even when we cannot explicitly extract the source of bias within each video frame or enumerate types of bias. Finally, to precisely evaluate static bias, we synthesize two new benchmarks, SCUBA for static cues in the background, and SCUFO for static cues in the foreground. With extensive experiments, we demonstrate that StillMix mitigates both types of static bias and improves video representations for downstream applications. Code is available at https://github.com/lihaoxin05/StillMix.
In this report, we present the ReLER@ZJU1 submission to the Ego4D Moment Queries Challenge in ECCV 2022. In this task, the goal is to retrieve and localize all instances of possible activities in egocentric videos. Ego4D dataset is challenging for the temporal action localization task as the temporal duration of the videos is quite long and each video contains multiple action instances with fine-grained action classes. To address these problems, we utilize a multi-scale transformer to classify different action categories and predict the boundary of each instance. Moreover, in order to better capture the long-term temporal dependencies in the long videos, we propose a segment-level recurrence mechanism. Compared with directly feeding all video features to the transformer encoder, the proposed segment-level recurrence mechanism alleviates the optimization difficulties and achieves better performance. The final submission achieved Recall@1,tIoU=0.5 score of 37.24, average mAP score of 17.67 and took 3-rd place on the leaderboard.
This report describes our submission to the Ego4D Moment Queries Challenge 2022. Our submission builds on ActionFormer, the state-of-the-art backbone for temporal action localization, and a trio of strong video features from SlowFast, Omnivore and EgoVLP. Our solution is ranked 2nd on the public leaderboard with 21.76% average mAP on the test set, which is nearly three times higher than the official baseline. Further, we obtain 42.54% Recall@1x at tIoU=0.5 on the test set, outperforming the top-ranked solution by a significant margin of 1.41 absolute percentage points. Our code is available at https://github.com/happyharrycn/actionformer_release.
This report describes Badgers@UW-Madison, our submission to the Ego4D Natural Language Queries (NLQ) Challenge. Our solution inherits the point-based event representation from our prior work on temporal action localization, and develops a Transformer-based model for video grounding. Further, our solution integrates several strong video features including SlowFast, Omnivore and EgoVLP. Without bells and whistles, our submission based on a single model achieves 12.64% Mean R@1 and is ranked 2nd on the public leaderboard. Meanwhile, our method garners 28.45% (18.03%) R@5 at tIoU=0.3 (0.5), surpassing the top-ranked solution by up to 5.5 absolute percentage points.
Human action recognition has drawn a lot of attention in the recent years due to the research and application significance. Most existing works on action recognition focus on learning effective spatial-temporal features from videos, but neglect the strong causal relationship among the precondition, action and effect. Such relationships are also crucial to the accuracy of action recognition. In this paper, we propose to model the causal relationships based on the precondition and effect to improve the performance of action recognition. Specifically, a Cycle-Reasoning model is proposed to capture the causal relationships for action recognition. To this end, we annotate precondition and effect for a large-scale action dataset. Experimental results show that the proposed Cycle-Reasoning model can effectively reason about the precondition and effect and can enhance action recognition performance.
We launch EVA, a vision-centric foundation model to explore the limits of visual representation at scale using only publicly accessible data. EVA is a vanilla ViT pre-trained to reconstruct the masked out image-text aligned vision features conditioned on visible image patches. Via this pretext task, we can efficiently scale up EVA to one billion parameters, and sets new records on a broad range of representative vision downstream tasks, such as image recognition, video action recognition, object detection, instance segmentation and semantic segmentation without heavy supervised training. Moreover, we observe quantitative changes in scaling EVA result in qualitative changes in transfer learning performance that are not present in other models. For instance, EVA takes a great leap in the challenging large vocabulary instance segmentation task: our model achieves almost the same state-of-the-art performance on LVISv1.0 dataset with over a thousand categories and COCO dataset with only eighty categories. Beyond a pure vision encoder, EVA can also serve as a vision-centric, multi-modal pivot to connect images and text. We find initializing the vision tower of a giant CLIP from EVA can greatly stabilize the training and outperform the training from scratch counterpart with much fewer samples and less compute, providing a new direction for scaling up and accelerating the costly training of multi-modal foundation models. To facilitate future research, we release all the code and models at https://github.com/baaivision/EVA.
Temporal Action Localization (TAL) methods typically operate on top of feature sequences from a frozen snippet encoder that is pretrained with the Trimmed Action Classification (TAC) tasks, resulting in a task discrepancy problem. While existing TAL methods mitigate this issue either by retraining the encoder with a pretext task or by end-to-end fine-tuning, they commonly require an overload of high memory and computation. In this work, we introduce Soft-Landing (SoLa) strategy, an efficient yet effective framework to bridge the transferability gap between the pretrained encoder and the downstream tasks by incorporating a light-weight neural network, i.e., a SoLa module, on top of the frozen encoder. We also propose an unsupervised training scheme for the SoLa module; it learns with inter-frame Similarity Matching that uses the frame interval as its supervisory signal, eliminating the need for temporal annotations. Experimental evaluation on various benchmarks for downstream TAL tasks shows that our method effectively alleviates the task discrepancy problem with remarkable computational efficiency.
Online Temporal Action Localization (On-TAL) aims to immediately provide action instances from untrimmed streaming videos. The model is not allowed to utilize future frames and any processing techniques to modify past predictions, making On-TAL much more challenging. In this paper, we propose a simple yet effective framework, termed SimOn, that learns to predict action instances using the popular Transformer architecture in an end-to-end manner. Specifically, the model takes the current frame feature as a query and a set of past context information as keys and values of the Transformer. Different from the prior work that uses a set of outputs of the model as past contexts, we leverage the past visual context and the learnable context embedding for the current query. Experimental results on the THUMOS14 and ActivityNet1.3 datasets show that our model remarkably outperforms the previous methods, achieving a new state-of-the-art On-TAL performance. In addition, the evaluation for Online Detection of Action Start (ODAS) demonstrates the effectiveness and robustness of our method in the online setting. The code is available at https://github.com/TuanTNG/SimOn
Young children are at an increased risk of contracting contagious diseases such as COVID-19 due to improper hand hygiene. An autonomous social agent that observes children while handwashing and encourages good hand washing practices could provide an opportunity for handwashing behavior to become a habit. In this article, we present a human action recognition system, which is part of the vision system of a social robot platform, to assist children in developing a correct handwashing technique. A modified convolution neural network (CNN) architecture with Channel Spatial Attention Bilinear Pooling (CSAB) frame, with a VGG-16 architecture as the backbone is trained and validated on an augmented dataset. The modified architecture generalizes well with an accuracy of 90% for the WHO-prescribed handwashing steps even in an unseen environment. Our findings indicate that the approach can recognize even subtle hand movements in the video and can be used for gesture detection and classification in social robotics.
Actions are about how we interact with the environment, including other people, objects, and ourselves. In this paper, we propose a novel multi-modal Holistic Interaction Transformer Network (HIT) that leverages the largely ignored, but critical hand and pose information essential to most human actions. The proposed "HIT" network is a comprehensive bi-modal framework that comprises an RGB stream and a pose stream. Each of them separately models person, object, and hand interactions. Within each sub-network, an Intra-Modality Aggregation module (IMA) is introduced that selectively merges individual interaction units. The resulting features from each modality are then glued using an Attentive Fusion Mechanism (AFM). Finally, we extract cues from the temporal context to better classify the occurring actions using cached memory. Our method significantly outperforms previous approaches on the J-HMDB, UCF101-24, and MultiSports datasets. We also achieve competitive results on AVA. The code will be available at https://github.com/joslefaure/HIT.
Traditional temporal action detection (TAD) usually handles untrimmed videos with small number of action instances from a single label (e.g., ActivityNet, THUMOS). However, this setting might be unrealistic as different classes of actions often co-occur in practice. In this paper, we focus on the task of multi-label temporal action detection that aims to localize all action instances from a multi-label untrimmed video. Multi-label TAD is more challenging as it requires for fine-grained class discrimination within a single video and precise localization of the co-occurring instances. To mitigate this issue, we extend the sparse query-based detection paradigm from the traditional TAD and propose the multi-label TAD framework of PointTAD. Specifically, our PointTAD introduces a small set of learnable query points to represent the important frames of each action instance. This point-based representation provides a flexible mechanism to localize the discriminative frames at boundaries and as well the important frames inside the action. Moreover, we perform the action decoding process with the Multi-level Interactive Module to capture both point-level and instance-level action semantics. Finally, our PointTAD employs an end-to-end trainable framework simply based on RGB input for easy deployment. We evaluate our proposed method on two popular benchmarks and introduce the new metric of detection-mAP for multi-label TAD. Our model outperforms all previous methods by a large margin under the detection-mAP metric, and also achieves promising results under the segmentation-mAP metric. Code is available at https://github.com/MCG-NJU/PointTAD.
In this paper, we study the transferability of ImageNet spatial and Kinetics spatio-temporal representations to multi-label Movie Trailer Genre Classification (MTGC). In particular, we present an extensive evaluation of the transferability of ConvNet and Transformer models pretrained on ImageNet and Kinetics to Trailers12k, a new manually-curated movie trailer dataset composed of 12,000 videos labeled with 10 different genres and associated metadata. We analyze different aspects that can influence transferability, such as frame rate, input video extension, and spatio-temporal modeling. In order to reduce the spatio-temporal structure gap between ImageNet/Kinetics and Trailers12k, we propose Dual Image and Video Transformer Architecture (DIViTA), which performs shot detection so as to segment the trailer into highly correlated clips, providing a more cohesive input for pretrained backbones and improving transferability (a 1.83% increase for ImageNet and 3.75% for Kinetics). Our results demonstrate that representations learned on either ImageNet or Kinetics are comparatively transferable to Trailers12k. Moreover, both datasets provide complementary information that can be combined to improve classification performance (a 2.91% gain compared to the top single pretraining). Interestingly, using lightweight ConvNets as pretrained backbones resulted in only a 3.46% drop in classification performance compared with the top Transformer while requiring only 11.82% of its parameters and 0.81% of its FLOPS.
When applied to autonomous vehicle (AV) settings, action recognition can enhance an environment model's situational awareness. This is especially prevalent in scenarios where traditional geometric descriptions and heuristics in AVs are insufficient. However, action recognition has traditionally been studied for humans, and its limited adaptability to noisy, un-clipped, un-pampered, raw RGB data has limited its application in other fields. To push for the advancement and adoption of action recognition into AVs, this work proposes a novel two-stage action recognition system, termed RALACs. RALACs formulates the problem of action recognition for road scenes, and bridges the gap between it and the established field of human action recognition. This work shows how attention layers can be useful for encoding the relations across agents, and stresses how such a scheme can be class-agnostic. Furthermore, to address the dynamic nature of agents on the road, RALACs constructs a novel approach to adapting Region of Interest (ROI) Alignment to agent tracks for downstream action classification. Finally, our scheme also considers the problem of active agent detection, and utilizes a novel application of fusing optical flow maps to discern relevant agents in a road scene. We show that our proposed scheme can outperform the baseline on the ICCV2021 Road Challenge dataset and by deploying it on a real vehicle platform, we provide preliminary insight to the usefulness of action recognition in decision making.
We study the task of robust feature representations, aiming to generalize well on multiple datasets for action recognition. We build our method on Transformers for its efficacy. Although we have witnessed great progress for video action recognition in the past decade, it remains challenging yet valuable how to train a single model that can perform well across multiple datasets. Here, we propose a novel multi-dataset training paradigm, MultiTrain, with the design of two new loss terms, namely informative loss and projection loss, aiming to learn robust representations for action recognition. In particular, the informative loss maximizes the expressiveness of the feature embedding while the projection loss for each dataset mines the intrinsic relations between classes across datasets. We verify the effectiveness of our method on five challenging datasets, Kinetics-400, Kinetics-700, Moments-in-Time, Activitynet and Something-something-v2 datasets. Extensive experimental results show that our method can consistently improve state-of-the-art performance. Code and models are released.
Skeleton-based human action recognition is a longstanding challenge due to its complex dynamics. Some fine-grain details of the dynamics play a vital role in classification. The existing work largely focuses on designing incremental neural networks with more complicated adjacent matrices to capture the details of joints relationships. However, they still have difficulties distinguishing actions that have broadly similar motion patterns but belong to different categories. Interestingly, we found that the subtle differences in motion patterns can be significantly amplified and become easy for audience to distinct through specified view directions, where this property haven't been fully explored before. Drastically different from previous work, we boost the performance by proposing a conceptually simple yet effective Multi-view strategy that recognizes actions from a collection of dynamic view features. Specifically, we design a novel Skeleton-Anchor Proposal (SAP) module which contains a Multi-head structure to learn a set of views. For feature learning of different views, we introduce a novel Angle Representation to transform the actions under different views and feed the transformations into the baseline model. Our module can work seamlessly with the existing action classification model. Incorporated with baseline models, our SAP module exhibits clear performance gains on many challenging benchmarks. Moreover, comprehensive experiments show that our model consistently beats down the state-of-the-art and remains effective and robust especially when dealing with corrupted data. Related code will be available on https://github.com/ideal-idea/SAP .