M2-RAAP: A Multi-Modal Recipe for Advancing Adaptation-based Pre-training towards Effective and Efficient Zero-shot Video-text Retrieval

We present a Multi-Modal Recipe for Advancing Adaptation-based Pre-training towards effective and efficient zero-shot video-text retrieval, dubbed M2-RAAP. Upon popular image-text models like CLIP, most current adaptation-based video-text pre-training methods are confronted by three major issues, i.e., noisy data corpus, time-consuming pre-training, and limited performance gain. Towards this end, we conduct a comprehensive study including four critical steps in video-text pre-training. Specifically, we investigate 1) data filtering and refinement, 2) video input type selection, 3) temporal modeling, and 4) video feature enhancement. We then summarize this empirical study into the M2-RAAP recipe, where our technical contributions lie in 1) the data filtering and text re-writing pipeline resulting in 1M high-quality bilingual video-text pairs, 2) the replacement of video inputs with key-frames to accelerate pre-training, and 3) the Auxiliary-Caption-Guided (ACG) strategy to enhance video features. We conduct extensive experiments by adapting three image-text foundation models on two refined video-text datasets from different languages, validating the robustness and reproducibility of M2-RAAP for adaptation-based pre-training. Results demonstrate that M2-RAAP yields superior performance with significantly reduced data (-90%) and time consumption (-95%), establishing a new SOTA on four English zero-shot retrieval datasets and two Chinese ones. We are preparing our refined bilingual data annotations and codebase, which will be available at https://github.com/alipay/Ant-Multi-Modal-Framework/tree/main/prj/M2_RAAP.

SOAR: Scene-debiasing Open-set Action Recognition

Deep learning models have a risk of utilizing spurious clues to make predictions, such as recognizing actions based on the background scene. This issue can severely degrade the open-set action recognition performance when the testing samples have different scene distributions from the training samples. To mitigate this problem, we propose a novel method, called Scene-debiasing Open-set Action Recognition (SOAR), which features an adversarial scene reconstruction module and an adaptive adversarial scene classification module. The former prevents the decoder from reconstructing the video background given video features, and thus helps reduce the background information in feature learning. The latter aims to confuse scene type classification given video features, with a specific emphasis on the action foreground, and helps to learn scene-invariant information. In addition, we design an experiment to quantify the scene bias. The results indicate that the current open-set action recognizers are biased toward the scene, and our proposed SOAR method better mitigates such bias. Furthermore, our extensive experiments demonstrate that our method outperforms state-of-the-art methods, and the ablation studies confirm the effectiveness of our proposed modules.

EVE: Efficient zero-shot text-based Video Editing with Depth Map Guidance and Temporal Consistency Constraints

Motivated by the superior performance of image diffusion models, more and more researchers strive to extend these models to the text-based video editing task. Nevertheless, current video editing tasks mainly suffer from the dilemma between the high fine-tuning cost and the limited generation capacity. Compared with images, we conjecture that videos necessitate more constraints to preserve the temporal consistency during editing. Towards this end, we propose EVE, a robust and efficient zero-shot video editing method. Under the guidance of depth maps and temporal consistency constraints, EVE derives satisfactory video editing results with an affordable computational and time cost. Moreover, recognizing the absence of a publicly available video editing dataset for fair comparisons, we construct a new benchmark ZVE-50 dataset. Through comprehensive experimentation, we validate that EVE could achieve a satisfactory trade-off between performance and efficiency. We will release our dataset and codebase to facilitate future researchers.

Generalized Relation Modeling for Transformer Tracking

Compared with previous two-stream trackers, the recent one-stream tracking pipeline, which allows earlier interaction between the template and search region, has achieved a remarkable performance gain. However, existing one-stream trackers always let the template interact with all parts inside the search region throughout all the encoder layers. This could potentially lead to target-background confusion when the extracted feature representations are not sufficiently discriminative. To alleviate this issue, we propose a generalized relation modeling method based on adaptive token division. The proposed method is a generalized formulation of attention-based relation modeling for Transformer tracking, which inherits the merits of both previous two-stream and one-stream pipelines whilst enabling more flexible relation modeling by selecting appropriate search tokens to interact with template tokens. An attention masking strategy and the Gumbel-Softmax technique are introduced to facilitate the parallel computation and end-to-end learning of the token division module. Extensive experiments show that our method is superior to the two-stream and one-stream pipelines and achieves state-of-the-art performance on six challenging benchmarks with a real-time running speed.

AiATrack: Attention in Attention for Transformer Visual Tracking

Transformer trackers have achieved impressive advancements recently, where the attention mechanism plays an important role. However, the independent correlation computation in the attention mechanism could result in noisy and ambiguous attention weights, which inhibits further performance improvement. To address this issue, we propose an attention in attention (AiA) module, which enhances appropriate correlations and suppresses erroneous ones by seeking consensus among all correlation vectors. Our AiA module can be readily applied to both self-attention blocks and cross-attention blocks to facilitate feature aggregation and information propagation for visual tracking. Moreover, we propose a streamlined Transformer tracking framework, dubbed AiATrack, by introducing efficient feature reuse and target-background embeddings to make full use of temporal references. Experiments show that our tracker achieves state-of-the-art performance on six tracking benchmarks while running at a real-time speed.

Distilling Inter-Class Distance for Semantic Segmentation

Knowledge distillation is widely adopted in semantic segmentation to reduce the computation cost.The previous knowledge distillation methods for semantic segmentation focus on pixel-wise feature alignment and intra-class feature variation distillation, neglecting to transfer the knowledge of the inter-class distance in the feature space, which is important for semantic segmentation. To address this issue, we propose an Inter-class Distance Distillation (IDD) method to transfer the inter-class distance in the feature space from the teacher network to the student network. Furthermore, semantic segmentation is a position-dependent task,thus we exploit a position information distillation module to help the student network encode more position information. Extensive experiments on three popular datasets: Cityscapes, Pascal VOC and ADE20K show that our method is helpful to improve the accuracy of semantic segmentation models and achieves the state-of-the-art performance. E.g. it boosts the benchmark model("PSPNet+ResNet18") by 7.50% in accuracy on the Cityscapes dataset.

Learning Dynamics via Graph Neural Networks for Human Pose Estimation and Tracking

Multi-person pose estimation and tracking serve as crucial steps for video understanding. Most state-of-the-art approaches rely on first estimating poses in each frame and only then implementing data association and refinement. Despite the promising results achieved, such a strategy is inevitably prone to missed detections especially in heavily-cluttered scenes, since this tracking-by-detection paradigm is, by nature, largely dependent on visual evidences that are absent in the case of occlusion. In this paper, we propose a novel online approach to learning the pose dynamics, which are independent of pose detections in current fame, and hence may serve as a robust estimation even in challenging scenarios including occlusion. Specifically, we derive this prediction of dynamics through a graph neural network~(GNN) that explicitly accounts for both spatial-temporal and visual information. It takes as input the historical pose tracklets and directly predicts the corresponding poses in the following frame for each tracklet. The predicted poses will then be aggregated with the detected poses, if any, at the same frame so as to produce the final pose, potentially recovering the occluded joints missed by the estimator. Experiments on PoseTrack 2017 and PoseTrack 2018 datasets demonstrate that the proposed method achieves results superior to the state of the art on both human pose estimation and tracking tasks.

Actor-Action Semantic Segmentation with Region Masks

In this paper, we study the actor-action semantic segmentation problem, which requires joint labeling of both actor and action categories in video frames. One major challenge for this task is that when an actor performs an action, different body parts of the actor provide different types of cues for the action category and may receive inconsistent action labeling when they are labeled independently. To address this issue, we propose an end-to-end region-based actor-action segmentation approach which relies on region masks from an instance segmentation algorithm. Our main novelty is to avoid labeling pixels in a region mask independently - instead we assign a single action label to these pixels to achieve consistent action labeling. When a pixel belongs to multiple region masks, max pooling is applied to resolve labeling conflicts. Our approach uses a two-stream network as the front-end (which learns features capturing both appearance and motion information), and uses two region-based segmentation networks as the back-end (which takes the fused features from the two-stream network as the input and predicts actor-action labeling). Experiments on the A2D dataset demonstrate that both the region-based segmentation strategy and the fused features from the two-stream network contribute to the performance improvements. The proposed approach outperforms the state-of-the-art results by more than 8% in mean class accuracy, and more than 5% in mean class IOU, which validates its effectiveness.