The landscape of video recognition has evolved significantly, shifting from traditional Convolutional Neural Networks (CNNs) to Transformer-based architectures for improved accuracy. While 3D CNNs have been effective at capturing spatiotemporal dynamics, recent Transformer models leverage self-attention to model long-range spatial and temporal dependencies. Despite achieving state-of-the-art performance on major benchmarks, Transformers remain computationally expensive, particularly with dense video data. To address this, we propose a lightweight Video Focal Modulation Network, DVFL-Net, which distills spatiotemporal knowledge from a large pre-trained teacher into a compact nano student model, enabling efficient on-device deployment. DVFL-Net utilizes knowledge distillation and spatial-temporal feature modulation to significantly reduce computation while preserving high recognition performance. We employ forward Kullback-Leibler (KL) divergence alongside spatio-temporal focal modulation to effectively transfer both local and global context from the Video-FocalNet Base (teacher) to the proposed VFL-Net (student). We evaluate DVFL-Net on UCF50, UCF101, HMDB51, SSV2, and Kinetics-400, benchmarking it against recent state-of-the-art methods in Human Action Recognition (HAR). Additionally, we conduct a detailed ablation study analyzing the impact of forward KL divergence. The results confirm the superiority of DVFL-Net in achieving an optimal balance between performance and efficiency, demonstrating lower memory usage, reduced GFLOPs, and strong accuracy, making it a practical solution for real-time HAR applications.
Online temporal action segmentation shows a strong potential to facilitate many HRI tasks where extended human action sequences must be tracked and understood in real time. Traditional action segmentation approaches, however, operate in an offline two stage approach, relying on computationally expensive video wide features for segmentation, rendering them unsuitable for online HRI applications. In order to facilitate online action segmentation on a stream of incoming video data, we introduce two methods for improved training and inference of backbone action recognition models, allowing them to be deployed directly for online frame level classification. Firstly, we introduce surround dense sampling whilst training to facilitate training vs. inference clip matching and improve segment boundary predictions. Secondly, we introduce an Online Temporally Aware Label Cleaning (O-TALC) strategy to explicitly reduce oversegmentation during online inference. As our methods are backbone invariant, they can be deployed with computationally efficient spatio-temporal action recognition models capable of operating in real time with a small segmentation latency. We show our method outperforms similar online action segmentation work as well as matches the performance of many offline models with access to full temporal resolution when operating on challenging fine-grained datasets.
Spatio-temporal action recognition has been a challenging task that involves detecting where and when actions occur. Current state-of-the-art action detectors are mostly anchor-based, requiring sensitive anchor designs and huge computations due to calculating large numbers of anchor boxes. Motivated by nascent anchor-free approaches, we propose Point3D, a flexible and computationally efficient network with high precision for spatio-temporal action recognition. Our Point3D consists of a Point Head for action localization and a 3D Head for action classification. Firstly, Point Head is used to track center points and knot key points of humans to localize the bounding box of an action. These location features are then piped into a time-wise attention to learn long-range dependencies across frames. The 3D Head is later deployed for the final action classification. Our Point3D achieves state-of-the-art performance on the JHMDB, UCF101-24, and AVA benchmarks in terms of frame-mAP and video-mAP. Comprehensive ablation studies also demonstrate the effectiveness of each module proposed in our Point3D.
There has been an increasing interest in spiking neural networks in recent years. SNNs are seen as hypothetical solutions for the bottlenecks of ANNs in pattern recognition, such as energy efficiency. But current methods such as ANN-to-SNN conversion and back-propagation do not take full advantage of these networks, and unsupervised methods have not yet reached a success comparable to advanced artificial neural networks. It is important to study the behavior of SNNs trained with unsupervised learning methods such as spike-timing dependent plasticity (STDP) on video classification tasks, including mechanisms to model motion information using spikes, as this information is critical for video understanding. This paper presents multiple methods of transposing temporal information into a static format, and then transforming the visual information into spikes using latency coding. These methods are paired with two types of temporal fusion known as early and late fusion, and are used to help the spiking neural network in capturing the spatio-temporal features from videos. In this paper, we rely on the network architecture of a convolutional spiking neural network trained with STDP, and we test the performance of this network when challenged with action recognition tasks. Understanding how a spiking neural network responds to different methods of movement extraction and representation can help reduce the performance gap between SNNs and ANNs. In this paper we show the effect of the similarity in the shape and speed of certain actions on action recognition with spiking neural networks, we also highlight the effectiveness of some methods compared to others.
Detecting and recognizing human action in videos with crowded scenes is a challenging problem due to the complex environment and diversity events. Prior works always fail to deal with this problem in two aspects: (1) lacking utilizing information of the scenes; (2) lacking training data in the crowd and complex scenes. In this paper, we focus on improving spatio-temporal action recognition by fully-utilizing the information of scenes and collecting new data. A top-down strategy is used to overcome the limitations. Specifically, we adopt a strong human detector to detect the spatial location of each frame. We then apply action recognition models to learn the spatio-temporal information from video frames on both the HIE dataset and new data with diverse scenes from the internet, which can improve the generalization ability of our model. Besides, the scenes information is extracted by the semantic segmentation model to assistant the process. As a result, our method achieved an average 26.05 wf\_mAP (ranking 1st place in the ACM MM grand challenge 2020: Human in Events).
Despite the recent advances in video classification, progress in spatio-temporal action recognition has lagged behind. A major contributing factor has been the prohibitive cost of annotating videos frame-by-frame. In this paper, we present a spatio-temporal action recognition model that is trained with only video-level labels, which are significantly easier to annotate. Our method leverages per-frame person detectors which have been trained on large image datasets within a Multiple Instance Learning framework. We show how we can apply our method in cases where the standard Multiple Instance Learning assumption, that each bag contains at least one instance with the specified label, is invalid using a novel probabilistic variant of MIL where we estimate the uncertainty of each prediction. Furthermore, we report the first weakly-supervised results on the AVA dataset and state-of-the-art results among weakly-supervised methods on UCF101-24.
Manipulation actions transform objects from an initial state into a final state. In this paper, we report on the use of object state transitions as a mean for recognizing manipulation actions. Our method is inspired by the intuition that object states are visually more apparent than actions from a still frame and thus provide information that is complementary to spatio-temporal action recognition. We start by defining a state transition matrix that maps action labels into a pre-state and a post-state. From each keyframe, we learn appearance models of objects and their states. Manipulation actions can then be recognized from the state transition matrix. We report results on the EPIC kitchen action recognition challenge.
The task of action recognition or action detection involves analyzing videos and determining what action or motion is being performed. The primary subject of these videos are predominantly humans performing some action. However, this requirement can be relaxed to generalize over other subjects such as animals or robots. The applications can range from anywhere between human-computer inter-action to automated video editing proposals. When we consider spatiotemporal action recognition, we deal with action localization. This task not only involves determining what action is being performed but also when and where itis being performed in said video. This paper aims to survey the plethora of approaches and algorithms attempted to solve this task, give a comprehensive comparison between them, explore various datasets available for the problem, and determine the most promising approaches.
This paper introduces a video dataset of spatio-temporally localized Atomic Visual Actions (AVA). The AVA dataset densely annotates 80 atomic visual actions in 430 15-minute video clips, where actions are localized in space and time, resulting in 1.58M action labels with multiple labels per person occurring frequently. The key characteristics of our dataset are: (1) the definition of atomic visual actions, rather than composite actions; (2) precise spatio-temporal annotations with possibly multiple annotations for each person; (3) exhaustive annotation of these atomic actions over 15-minute video clips; (4) people temporally linked across consecutive segments; and (5) using movies to gather a varied set of action representations. This departs from existing datasets for spatio-temporal action recognition, which typically provide sparse annotations for composite actions in short video clips. We will release the dataset publicly. AVA, with its realistic scene and action complexity, exposes the intrinsic difficulty of action recognition. To benchmark this, we present a novel approach for action localization that builds upon the current state-of-the-art methods, and demonstrates better performance on JHMDB and UCF101-24 categories. While setting a new state of the art on existing datasets, the overall results on AVA are low at 15.6% mAP, underscoring the need for developing new approaches for video understanding.