UniFormer: Unified Transformer for Efficient Spatiotemporal Representation Learning
Shenzhen Institute of Advanced Technology, Chinese Academy of Sciences（中国科学院深圳先进技术研究院）；University of Chinese Academy of Sciences（中国科学院大学）；Shanghai AI Laboratory（上海人工智能实验室）；SenseTime Research（商汤科技研究院）；The Chinese University of Hong Kong（香港中文大学）
It is a challenging task to learn rich and multi-scale spatiotemporal semantics from high-dimensional videos, due to large local redundancy and complex global dependency between video frames. The recent advances in this research have been mainly driven by 3D convolutional neural networks and vision transformers. Although 3D convolution can efficiently aggregate local context to suppress local redundancy from a small 3D neighborhood, it lacks the capability to capture global dependency because of the limited receptive field. Alternatively, vision transformers can effectively capture long-range dependency by self-attention mechanism, while having the limitation on reducing local redundancy with blind similarity comparison among all the tokens in each layer. Based on these observations, we propose a novel Unified transFormer (UniFormer) which seamlessly integrates merits of 3D convolution and spatiotemporal self-attention in a concise transformer format, and achieves a preferable balance between computation and accuracy. Different from traditional transformers, our relation aggregator can tackle both spatiotemporal redundancy and dependency, by learning local and global token affinity respectively in shallow and deep layers. We conduct extensive experiments on the popular video benchmarks, e.g., Kinetics-400, Kinetics-600, and Something-Something V1&V2. With only ImageNet-1K pretraining, our UniFormer achieves 82.9%/84.8% top-1 accuracy on Kinetics-400/Kinetics-600, while requiring 10x fewer GFLOPs than other state-of-the-art methods. For Something-Something V1 and V2, our UniFormer achieves new state-of-the-art performances of 60.9% and 71.2% top-1 accuracy respectively. Code is available at https://github.com/Sense-X/UniFormer.
Shenzhen Institute of Advanced Technology, Chinese Academy of Sciences（中国科学院深圳先进技术研究院）；University of Chinese Academy of Sciences（中国科学院大学）；Shanghai AI Laboratory（上海人工智能实验室）；The University of Hong Kong（香港大学）；State Key Laboratory for Novel Software Technology, Nanjing University（南京大学计算机软件新技术国家重点实验室）
Learning discriminative spatiotemporal representation is the key problem of video understanding. Recently, Vision Transformers (ViTs) have shown their power in learning long-term video dependency with self-attention. Unfortunately, they exhibit limitations in tackling local video redundancy, due to the blind global comparison among tokens. UniFormer has successfully alleviated this issue, by unifying convolution and self-attention as a relation aggregator in the transformer format. However, this model has to require a tiresome and complicated image-pretraining phrase, before being finetuned on…
UniFormer: Unifying Convolution and Self-attention for Visual Recognition
Shenzhen Institute of Advanced Technology, Chinese Academy of Sciences（中国科学院深圳先进技术研究院）；University of Chinese Academy of Sciences（中国科学院大学）；Shanghai Artificial Intelligence Laboratory（上海人工智能实验室）；National University of Singapore（新加坡国立大学）；SenseTime Research（商汤科技研究院）；the Chinese University of Hong Kong（香港中文大学）
University of Toronto（多伦多大学）；University of Guelph（圭尔夫大学）；Université de Lyon（里昂大学）；INSA-Lyon（里昂国立应用科学学院）；Modiface Inc（Modiface公司）；Vector Institute（矢量研究所）
In this paper we introduce a Transformer-based approach to video object segmentation (VOS). To address compounding error and scalability issues of prior work, we propose a scalable, end-to-end method for VOS called Sparse Spatiotemporal Transformers (SST). SST extracts per-pixel representations for each object in a video using sparse attention over spatiotemporal features. Our attention-based formulation for VOS allows a model to learn to attend over a history of multiple frames and provides suitable inductive bias for performing correspondence-like computations necessary for solving motion se…
Beijing Institute of Technology（北京理工大学）；Baidu Research（百度研究院）；Inception Institute of Artificial Intelligence（人工智能起源研究所）；National Engineering Laboratory of Deep Learning Technology and Application（深度学习技术及应用国家工程实验室）；University of Kentucky（肯塔基大学）
Existing LiDAR-based 3D object detectors usually focus on the single-frame detection, while ignoring the spatiotemporal information in consecutive point cloud frames. In this paper, we propose an end-to-end online 3D video object detector that operates on point cloud sequences. The proposed model comprises a spatial feature encoding component and a spatiotemporal feature aggregation component. In the former component, a novel Pillar Message Passing Network (PMPNet) is proposed to encode each discrete point cloud frame. It adaptively collects information for a pillar node from its neighbors by …
Beihang University（北京航空航天大学）；Tsinghua University（清华大学）；University of Science and Technology of China（中国科学技术大学）；Pengcheng Laboratory（鹏城实验室）；Xidian University（西安电子科技大学）
Recently, the Transformer module has been transplanted from natural language processing to computer vision. This paper applies the Transformer to video-based person re-identification, where the key issue is to extract the discriminative information from a tracklet. We show that, despite the strong learning ability, the vanilla Transformer suffers from an increased risk of over-fitting, arguably due to a large number of attention parameters and insufficient training data. To solve this problem, we propose a novel pipeline where the model is pre-trained on a set of synthesized video data and the…
Environment perception including detection, classification, tracking, and motion prediction are key enablers for automated driving systems and intelligent transportation applications. Fueled by the advances in sensing technologies and machine learning techniques, LiDAR-based sensing systems have become a promising solution. The current challenges of this solution are how to effectively combine different perception tasks into a single backbone and how to efficiently learn the spatiotemporal features directly from point cloud sequences. In this research, we propose a novel spatiotemporal attenti…
This paper considers the problem of spatiotemporal object-centric reasoning in videos. Central to our approach is the notion of object permanence, i.e., the ability to reason about the location of objects as they move through the video while being occluded, contained or carried by other objects. Existing deep learning based approaches often suffer from spatiotemporal biases when applied to video reasoning problems. We propose Hopper, which uses a Multi-hop Transformer for reasoning object permanence in videos. Given a video and a localization query, Hopper reasons over image and object tracks …
Earth Observatory is a growing research area that can capitalize on the powers of AI for short time forecasting, a Now-casting scenario. In this work, we tackle the challenge of weather forecasting using a video transformer network. Vision transformer architectures have been explored in various applications, with major constraints being the computational complexity of Attention and the data hungry training. To address these issues, we propose the use of Video Swin-Transformer, coupled with a dedicated augmentation scheme. Moreover, we employ gradual spatial reduction on the encoder side and cr…
Nanjing University（南京大学）；Shanghai AI Laboratory（上海人工智能实验室）；The University of Hong Kong（香港大学）
3D visual perception tasks, including 3D detection and map segmentation based on multi-camera images, are essential for autonomous driving systems. In this work, we present a new framework termed BEVFormer, which learns unified BEV representations with spatiotemporal transformers to support multiple autonomous driving perception tasks. In a nutshell, BEVFormer exploits both spatial and temporal information by interacting with spatial and temporal space through predefined grid-shaped BEV queries. To aggregate spatial information, we design spatial cross-attention that each BEV query extracts th…
Beijing Institute of Technology（北京理工大学）；Inception Institute of Artificial Intelligence（起源人工智能研究院）；King Abdullah University of Science and Technology（阿卜杜拉国王科技大学）；Indiana University（印第安纳大学）；University of Kentucky（肯塔基大学）
Previous works for LiDAR-based 3D object detection mainly focus on the single-frame paradigm. In this paper, we propose to detect 3D objects by exploiting temporal information in multiple frames, i.e., the point cloud videos. We empirically categorize the temporal information into short-term and long-term patterns. To encode the short-term data, we present a Grid Message Passing Network (GMPNet), which considers each grid (i.e., the grouped points) as a node and constructs a k-NN graph with the neighbor grids. To update features for a grid, GMPNet iteratively collects information from its neig…
Stony Brook University（石溪大学）；Mohamed bin Zayed University of AI（穆罕默德·本·扎耶德人工智能大学）；Australian National University（澳大利亚国立大学）；Linköping University（林雪平大学）
In this paper, we propose self-supervised training for video transformers using unlabeled video data. From a given video, we create local and global spatiotemporal views with varying spatial sizes and frame rates. Our self-supervised objective seeks to match the features of these different views representing the same video, to be invariant to spatiotemporal variations in actions. To the best of our knowledge, the proposed approach is the first to alleviate the dependency on negative samples or dedicated memory banks in Self-supervised Video Transformer (SVT). Further, owing to the flexibility …
Highly Efficient 3D Human Pose Tracking from Events with Spiking Spatiotemporal Transformer
Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences（深圳先进技术研究院，中国科学院）；Faculty of Computer Science and Control Engineering, Shenzhen University of Advanced Technology（计算机科学与控制工程学院，深圳大学）；Department of Electrical and Computer Engineering, University of Alberta（电气与计算机工程系，阿尔伯塔大学）
Mohamed Bin Zayed University of Artificial Intelligence（穆罕默德·本·扎耶德人工智能大学）
Learning spatiotemporal features is an important task for efficient video understanding especially in medical images such as echocardiograms. Convolutional neural networks (CNNs) and more recent vision transformers (ViTs) are the most commonly used methods with limitations per each. CNNs are good at capturing local context but fail to learn global information across video frames. On the other hand, vision transformers can incorporate global details and long sequences but are computationally expensive and typically require more data to train. In this paper, we propose a method that addresses th…
University of Science and Technology Beijing（北京科技大学）；Microsoft Research（微软研究院）；Shunde Graduate School of University of Science and Technology Beijing（北京科技大学顺德研究生院）
Compressed video super-resolution (VSR) aims to restore high-resolution frames from compressed low-resolution counterparts. Most recent VSR approaches often enhance an input frame by borrowing relevant textures from neighboring video frames. Although some progress has been made, there are grand challenges to effectively extract and transfer high-quality textures from compressed videos where most frames are usually highly degraded. In this paper, we propose a novel Frequency-Transformer for compressed video super-resolution (FTVSR) that conducts self-attention over a joint space-time-frequency …
Institute of Information Engineering, Chinese Academy of Science（中国科学院信息工程研究所）；School of Cyber Security, University of Chinese Academy of Sciences（中国科学院大学网络空间安全学院）；School of Communication and Information Engineering, Shanghai University（上海大学通信与信息工程学院）
While the abuse of deepfake technology has caused serious concerns recently, how to detect deepfake videos is still a challenge due to the high photo-realistic synthesis of each frame. Existing image-level approaches often focus on single frame and ignore the spatiotemporal cues hidden in deepfake videos, resulting in poor generalization and robustness. The key of a video-level detector is to fully exploit the spatiotemporal inconsistency distributed in local facial regions across different frames in deepfake videos. Inspired by that, this paper proposes a simple yet effective patch-level appr…
Phase-Specific Augmented Reality Guidance for Microscopic Cataract Surgery Using Long-Short Spatiotemporal Aggregation Transformer
Shanghai Jiao Tong University（上海交通大学）；Xinhua Hospital Affiliated to Shanghai Jiao Tong University School of Medicine（上海交通大学医学院附属新华医院）；Rice University（莱斯大学）；University of Münster（明斯特大学）
DualX-VSR: Dual Axial Spatial$\times$Temporal Transformer for Real-World Video Super-Resolution without Motion Compensation
University of Science and Technology of China（中国科学技术大学）；Shanghai Artificial Intelligence Laboratory（上海人工智能实验室）；Shanghai Jiao Tong University（上海交通大学）；Tsinghua University（清华大学）；Sun Yat-sen University（中山大学）；Shenzhen Institute of Advanced Technology, Chinese Academy of Sciences（中国科学院深圳先进技术研究院）
Beijing University of Posts and Telecommunications（北京邮电大学）
Exemplar-based video colorization is an essential technique for applications like old movie restoration. Although recent methods perform well in still scenes or scenes with regular movement, they always lack robustness in moving scenes due to their weak ability in modeling long-term dependency both spatially and temporally, leading to color fading, color discontinuity or other artifacts. To solve this problem, we propose an exemplar-based video colorization framework with long-term spatiotemporal dependency. To enhance the long-term spatial dependency, a parallelized CNN-Transformer block and …
Traffic forecasting is an important element of mobility management, an important key that drives the logistics industry. Over the years, lots of work have been done in Traffic forecasting using time series as well as spatiotemporal dynamic forecasting. In this paper, we explore the use of vision transformer in a UNet setting. We completely remove all convolution-based building blocks in UNet, while using 3D shifted window transformer in both encoder and decoder branches. In addition, we experiment with the use of feature mixing just before patch encoding to control the inter-relationship of th…
CardioDiT: Latent Diffusion Transformers for 4D Cardiac MRI Synthesis
Institute for Artificial Intelligence in Cardiovascular Medicine, Medical Faculty of Heidelberg University, Heidelberg University（海德堡大学医学院心血管医学人工智能研究所，海德堡大学）；Department of Cardiology, Angiology, Pneumology, Heidelberg University Hospital（海德堡大学医院心脏病学、血管学、肺病学系）；DZHK (German Centre for Cardiovascular Research), Partner Site Heidelberg/Mannheim（德国心血管研究中心（DZHK），海德堡/曼海姆合作站点）；Department of Radiology and Nuclear Medicine, University Medical Center Mannheim（曼海姆大学医学中心放射学与核医学系）；Section for Invasive Cardiology and Electrophysiology, I. Medical Department, Cardiology, Hemostasiology and Intensive Care, University Medical Centre Mannheim（曼海姆大学医学中心第一内科（心脏病学、止血学与重症监护）介入心脏病学与电生理学科）；University Medical Center of the Johannes Gutenberg-University Mainz（美因茨约翰内斯·古腾堡大学医学中心）；Department of Cardiology, Angiology, Hemostasis, and Medical Intensive Care, University Medical Centre Mannheim, Medical Faculty Mannheim, University of Heidelberg（海德堡大学曼海姆医学院曼海姆大学医学中心心脏病学、血管学、止血学与内科重症监护系）
本文提出CardioDiT，一种基于扩散变换器的4D潜在扩散模型，用于合成短轴 cine CMR，通过联合建模空间和时间实现连续心脏动态生成。
Institute of Information Engineering, Chinese Academy of Sciences（中国科学院信息工程研究所）
Graph convolutional networks (GCNs) have emerged as dominant methods for skeleton-based action recognition. However, they still suffer from two problems, namely, neighborhood constraints and entangled spatiotemporal feature representations. Most studies have focused on improving the design of graph topology to solve the first problem but they have yet to fully explore the latter. In this work, we design a disentangled spatiotemporal transformer (DSTT) block to overcome the above limitations of GCNs in three steps: (i) feature disentanglement for spatiotemporal decomposition;(ii) global spatiot…
Seoul National University（首尔大学）；SungKyunKwan University（成均馆大学）；Brookhaven National Lab（布鲁克海文国家实验室）
Modeling spatiotemporal brain dynamics from high-dimensional data, such as functional Magnetic Resonance Imaging (fMRI), is a formidable task in neuroscience. Existing approaches for fMRI analysis utilize hand-crafted features, but the process of feature extraction risks losing essential information in fMRI scans. To address this challenge, we present SwiFT (Swin 4D fMRI Transformer), a Swin Transformer architecture that can learn brain dynamics directly from fMRI volumes in a memory and computation-efficient manner. SwiFT achieves this by implementing a 4D window multi-head self-attention mec…
Zenseact AB（Zenseact公司）；Lund University（隆德大学）；Linköping University（林雪平大学）；Chalmers University of Technology（查尔姆斯理工大学）
We propose the task Future Object Detection, in which the goal is to predict the bounding boxes for all visible objects in a future video frame. While this task involves recognizing temporal and kinematic patterns, in addition to the semantic and geometric ones, it only requires annotations in the standard form for individual, single (future) frames, in contrast to expensive full sequence annotations. We propose to tackle this task with an end-to-end method, in which a detection transformer is trained to directly output the future objects. In order to make accurate predictions about the future…
A 20% rise in car crashes in 2021 compared to 2020 has been observed as a result of increased distraction and drowsiness. Drowsy and distracted driving are the cause of 45% of all car crashes. As a means to decrease drowsy and distracted driving, detection methods using computer vision can be designed to be low-cost, accurate, and minimally invasive. This work investigated the use of the vision transformer to outperform state-of-the-art accuracy from 3D-CNNs. Two separate transformers were trained for drowsiness and distractedness. The drowsy video transformer model was trained on the National…
Insights from Visual Cognition: Understanding Human Action Dynamics with Overall Glance and Refined Gaze Transformer
Department of Computational Engineering, Lappeenranta-Lahti University of Technology LUT, Finland（拉彭兰塔-拉赫蒂理工大学计算工程系，芬兰）；Faculty of Information Technology, Brno University of Technology, Czech Republic（布尔诺理工大学信息技术学院，捷克共和国）
The Hong Kong Polytechnic University（香港理工大学）；DAMO Academy, Alibaba（阿里巴巴达摩院）
Transformer-based methods have recently achieved great advancement on 2D image-based vision tasks. For 3D video-based tasks such as action recognition, however, directly applying spatiotemporal transformers on video data will bring heavy computation and memory burdens due to the largely increased number of patches and the quadratic complexity of self-attention computation. How to efficiently and effectively model the 3D self-attention of video data has been a great challenge for transformers. In this paper, we propose a Temporal Patch Shift (TPS) method for efficient 3D self-attention modeling…
Learning Natural Consistency Representation for Face Forgery Video Detection
Institute of Information Engineering, Chinese Academy of Sciences（中国科学院信息工程研究所）；University of Chinese Academy of Sciences（中国科学院大学）；Tsinghua University（清华大学）；RealAI
Computed tomography (CT) imaging could be very practical for diagnosing various diseases. However, the nature of the CT images is even more diverse since the resolution and number of the slices of a CT scan are determined by the machine and its settings. Conventional deep learning models are hard to tickle such diverse data since the essential requirement of the deep neural network is the consistent shape of the input data. In this paper, we propose a novel, effective, two-step-wise approach to tickle this issue for COVID-19 symptom classification thoroughly. First, the semantic feature embedd…
Google Research（谷歌研究院）；Michigan State University（密歇根州立大学）；Brown University（布朗大学）
Video understanding requires reasoning at multiple spatiotemporal resolutions -- from short fine-grained motions to events taking place over longer durations. Although transformer architectures have recently advanced the state-of-the-art, they have not explicitly modelled different spatiotemporal resolutions. To this end, we present Multiview Transformers for Video Recognition (MTV). Our model consists of separate encoders to represent different views of the input video with lateral connections to fuse information across views. We present thorough ablation studies of our model and show that MT…
University of Science and Technology Beijing（北京科技大学）；University of Sydney（悉尼大学）；Microsoft Research（微软研究院）
Video Super-Resolution (VSR) aims to restore high-resolution (HR) videos from low-resolution (LR) videos. Existing VSR techniques usually recover HR frames by extracting pertinent textures from nearby frames with known degradation processes. Despite significant progress, grand challenges are remained to effectively extract and transmit high-quality textures from high-degraded low-quality sequences, such as blur, additive noises, and compression artifacts. In this work, a novel Frequency-Transformer (FTVSR) is proposed for handling low-quality videos that carry out self-attention in a combined …
This paper studies a conceptually simple extension of Masked Autoencoders (MAE) to spatiotemporal representation learning from videos. We randomly mask out spacetime patches in videos and learn an autoencoder to reconstruct them in pixels. Interestingly, we show that our MAE method can learn strong representations with almost no inductive bias on spacetime (only except for patch and positional embeddings), and spacetime-agnostic random masking performs the best. We observe that the optimal masking ratio is as high as 90% (vs. 75% on images), supporting the hypothesis that this ratio is related…
The University of Tokyo（东京大学）；Wuhan University（武汉大学）；RIKEN（理化学研究所）；Microsoft Research Asia（微软亚洲研究院）
Convolutional neural networks (CNN) and Transformers have made impressive progress in the field of remote sensing change detection (CD). However, both architectures have inherent shortcomings: CNN are constrained by a limited receptive field that may hinder their ability to capture broader spatial contexts, while Transformers are computationally intensive, making them costly to train and deploy on large datasets. Recently, the Mamba architecture, based on state space models, has shown remarkable performance in a series of natural language processing tasks, which can effectively compensate for …
