The objective of this paper is visual-only self-supervised video representation learning. We make the following contributions: (i) we investigate the benefit of adding semantic-class positives to instance-based Info Noise Contrastive Estimation (InfoNCE) training, showing that this form of supervised contrastive learning leads to a clear improvement in performance; (ii) we propose a novel self-supervised co-training scheme to improve the popular infoNCE loss, exploiting the complementary information from different views, RGB streams and optical flow, of the same data source by using one view to obtain positive class samples for the other; (iii) we thoroughly evaluate the quality of the learnt representation on two different downstream tasks: action recognition and video retrieval. In both cases, the proposed approach demonstrates state-of-the-art or comparable performance with other self-supervised approaches, whilst being significantly more efficient to train, i.e. requiring far less training data to achieve similar performance.

We use the S3D architecture for all experiments. CoCLR), S3D is followed by a non-linear projection head. The projection head is removed when evaluating downstream tasks. The detailed dimensions are shown in Table 1.Stage Detail Output size: T HW C S3D followed by average pooling 1 1 When evaluating the pretrained representation for action classification, we replace the non-linear projection head with a single linear layer for the classification tasks. The history queue is used in all pretraining experiments (including both InfoNCE and CoCLR).

Recent advances in diffusion models have revolutionized video generation, offering superior temporal consistency and visual quality compared to traditional generative adversarial networks-based approaches. While this emerging field shows tremendous promise in applications, it faces significant challenges in motion consistency, computational efficiency, and ethical considerations. This survey provides a comprehensive review of diffusion-based video generation, examining its evolution, technical foundations, and practical applications. We present a systematic taxonomy of current methodologies, analyze architectural innovations and optimization strategies, and investigate applications across low-level vision tasks such as denoising and super-resolution. Additionally, we explore the synergies between diffusion-based video generation and related domains, including video representation learning, question answering, and retrieval. Compared to the existing surveys (Lei et al., 2024a;b; Melnik et al., 2024; Cao et al., 2023; Xing et al., 2024c) which focus on specific aspects of video generation, such as human video synthesis (Lei et al., 2024a) or long-form content generation (Lei et al., 2024b), our work provides a broader, more updated, and more fine-grained perspective on diffusion-based approaches with a special section for evaluation metrics, industry solutions, and training engineering techniques in video generation. This survey serves as a foundational resource for researchers and practitioners working at the intersection of diffusion models and video generation, providing insights into both the theoretical frameworks and practical implementations that drive this rapidly evolving field. A structured list of re-These authors contributed equally to this work.

Weaknesses: # Related Work I believe that the related work section has been poorly executed. It simply lists numerous papers from 4 domains of existing research. This provides limited information about the position of the proposed work w.r.t these existing works. A more detailed discussion of the contrast between the proposed work and existing literature, or the similarities, or the parts that have been directly adopted is generally expected from the related work section. Furthermore, the authors may have missed the following highly relevant papers which are also discussed in my Baselines section below: 1) Yan, Xueting, et al. "ClusterFit: Improving Generalization of Visual Representations."

The objective of this paper is visual-only self-supervised video representation learning. We make the following contributions: (i) we investigate the benefit of adding semantic-class positives to instance-based Info Noise Contrastive Estimation (InfoNCE) training, showing that this form of supervised contrastive learning leads to a clear improvement in performance; (ii) we propose a novel self-supervised co-training scheme to improve the popular infoNCE loss, exploiting the complementary information from different views, RGB streams and optical flow, of the same data source by using one view to obtain positive class samples for the other; (iii) we thoroughly evaluate the quality of the learnt representation on two different downstream tasks: action recognition and video retrieval. In both cases, the proposed approach demonstrates state-of-the-art or comparable performance with other self-supervised approaches, whilst being significantly more efficient to train, i.e. requiring far less training data to achieve similar performance.

Self-supervised video representation learning aimed at maximizing similarity between different temporal segments of one video, in order to enforce feature persistence over time. This leads to loss of pertinent information related to temporal relationships, rendering actions such as `enter' and `leave' to be indistinguishable. To mitigate this limitation, we propose Latent Time Navigation (LTN), a time-parameterized contrastive learning strategy that is streamlined to capture fine-grained motions. Specifically, we maximize the representation similarity between different video segments from one video, while maintaining their representations time-aware along a subspace of the latent representation code including an orthogonal basis to represent temporal changes. Our extensive experimental analysis suggests that learning video representations by LTN consistently improves performance of action classification in fine-grained and human-oriented tasks (e.g., on Toyota Smarthome dataset). In addition, we demonstrate that our proposed model, when pre-trained on Kinetics-400, generalizes well onto the unseen real world video benchmark datasets UCF101 and HMDB51, achieving state-of-the-art performance in action recognition.

In this paper, we propose a novel learning scheme for self-supervised video representation learning. Motivated by how humans understand videos, we propose to first learn general visual concepts then attend to discriminative local areas for video understanding. Specifically, we utilize static frame and frame difference to help decouple static and dynamic concepts, and respectively align the concept distributions in latent space. We add diversity and fidelity regularizations to guarantee that we learn a compact set of meaningful concepts. Then we employ a cross-attention mechanism to aggregate detailed local features of different concepts, and filter out redundant concepts with low activations to perform local concept contrast. Extensive experiments demonstrate that our method distills meaningful static and dynamic concepts to guide video understanding, and obtains state-of-the-art results on UCF-101, HMDB-51, and Diving-48.

Self-Supervised Learning has become an exciting direction in AI community. Predicting What You Already Know Helps: Provable Self-Supervised Learning. For self-supervised learning, Rationality implies generalization, provably. Can Pretext-Based Self-Supervised Learning Be Boosted by Downstream Data? FAIR Self-Supervision Benchmark [pdf] [repo]: various benchmark (and legacy) tasks for evaluating quality of visual representations learned by various self-supervision approaches.

Contrastive learning applied to self-supervised representation learning has seen a resurgence in deep models. In this paper, we find that existing contrastive learning based solutions for self-supervised video recognition focus on inter-variance encoding but ignore the intra-variance existing in clips within the same video. We thus propose to learn dual representations for each clip which (i) encode intra-variance through a shuffle-rank pretext task; (ii) encode inter-variance through a temporal coherent contrastive loss. Experiment results show that our method plays an essential role in balancing inter and intra variances and brings consistent performance gains on multiple backbones and contrastive learning frameworks. Integrated with SimCLR and pretrained on Kinetics-400, our method achieves 82.0% and 51.2% downstream classification accuracy on UCF101 and HMDB51 test sets respectively and 46.1% video retrieval accuracy on UCF101, outperforming both pretext-task based and contrastive learning based counterparts.

Deep neural networks based on Generative Adversarial Networks (GANs) have enabled end-to-end trainable photorealistic text-to-image generation. Researchers have also developed methods designed to increase user control over the process, such as dialogue-based methods that enable inputting instructions to designate the relative positions of objects in a generated scene. However, the language that can be used in these processes is restricted, and the generated images are limited to synthetic 3D visualizations or cartoons. A team from Google Research has targeted these text-to-image shortcomings with a new system called Tag-Retrieve-Compose Synthesize (TReCS), which exploits both user text and mouse traces. The method is proposed in the recent paper Text-to-Image Generation Grounded by Fine-Grained User Attention.