While existing query-based 3D end-to-end visual trackers integrate detection and tracking via the *tracking-by-attention* paradigm, these two chicken-and-egg tasks encounter optimization difficulties when sharing the same parameters. Our findings reveal that these difficulties arise due to two inherent constraints on the self-attention mechanism, i.e., over-deduplication for object queries and self-centric attention for track queries. In contrast, removing self-attention mechanism not only minimally impacts regression predictions of the tracker, but also tends to generate more latent candidate boxes. Based on these analyses, we present SynCL, a novel plug-and-play synergistic training strategy designed to co-facilitate multi-task learning for detection and tracking. Specifically, we propose a Task-specific Hybrid Matching module for a weight-shared cross-attention-based decoder that matches the targets of track queries with multiple object queries to exploit promising candidates overlooked by the self-attention mechanism and the bipartite matching. To flexibly select optimal candidates for the one-to-many matching, we also design a Dynamic Query Filtering module controlled by model training status. Moreover, we introduce Instance-aware Contrastive Learning to break through the barrier of self-centric attention for track queries, effectively bridging the gap between detection and tracking. Without additional inference costs, SynCL consistently delivers improvements in various benchmarks and achieves state-of-the-art performance with $58.9\%$ AMOTA on the nuScenes dataset. Code and raw results are available at <https://github.com/shubolin028/SynCL>.

The tracking-by-detection paradigm currently dominates multiple target tracking algorithms. It usually includes three tasks: target detection, appearance feature embedding, and data association. Carrying out these three tasks successively usually leads to lower tracking efficiency. In this paper, we propose a one-stage anchor-free multiple task learning framework which carries out target detection and appearance feature embedding in parallel to substantially increase the tracking speed. This framework simultaneously predicts a target detection and produces a feature embedding for each location, by sharing a pyramid of feature maps. We propose a deformable local attention module which utilizes the correlations between features at different locations within a target to obtain more discriminative features. We further propose a task-aware prediction module which utilizes deformable convolutions to select the most suitable locations for the different tasks. At the selected locations, classification of samples into foreground or background, appearance feature embedding, and target box regression are carried out. Two effective training strategies, regression range overlapping and sample reweighting, are proposed to reduce missed detections in dense scenes. Ambiguous samples whose identities are difficult to determine are effectively dealt with to obtain more accurate feature embedding of target appearance. An appearance-enhanced non-maximum suppression is proposed to reduce over-suppression of true targets in crowded scenes. Based on the one-stage anchor-free network with the deformable local attention module and the task-aware prediction module, we implement a new online multiple target tracker. Experimental results show that our tracker achieves a very fast speed while maintaining a high tracking accuracy.

Self-supervised learning on large-scale Vision Transformers (ViTs) as pre-training methods has achieved promising downstream performance. Yet, how much these pre-training paradigms promote lightweight ViTs' performance is considerably less studied. In this work, we develop and benchmark several self-supervised pre-training methods on image classification tasks and some downstream dense prediction tasks. We surprisingly find that if proper pre-training is adopted, even vanilla lightweight ViTs show comparable performance to previous SOTA networks with delicate architecture design. It breaks the recently popular conception that vanilla ViTs are not suitable for vision tasks in lightweight regimes. We also point out some defects of such pre-training, e.g., failing to benefit from large-scale pre-training data and showing inferior performance on data-insufficient downstream tasks. Furthermore, we analyze and clearly show the effect of such pre-training by analyzing the properties of the layer representation and attention maps for related models. Finally, based on the above analyses, a distillation strategy during pre-training is developed, which leads to further downstream performance improvement for MAE-based pre-training. Code is available at https://github.com/wangsr126/mae-lite.

Open- vocabulary object detection aims to detect novel object categories beyond the training set. The advanced open- vocabulary two-stage detectors employ instance-level visual-to- visual knowledge distillation to align the visual space of the detector with the semantic space of the Pre-trained Visual-Language Model (PVLM). However, in the more efficient one-stage detector, the absence of class-agnostic object proposals hinders the knowledge distil-lation on unseen objects, leading to severe performance degradation. In this paper, we propose a hierarchical visual-language knowledge distillation method, i.e., Hi-erKD, for open-vocabulary one-stage detection. Specifi-cally, a global-level knowledge distillation is explored to transfer the knowledge of unseen categories from the PVLM to the detector. Moreover, we combine the proposed global-level knowledge distillation and the common instance-level knowledge distillation to learn the knowledge of seen and unseen categories simultaneously. Extensive experiments on MS-COCO show that our method significantly surpasses the previous best one-stage detector with 11.9% and 6.7% AP50 gains under the zero-shot detection and generalized zero-shot detection settings, and reduces the AP50performance gap from 14% to 7.3% compared to the best two-stage detector. Code will be released at this url11https://qithub.com/menqqiDyanqqe/HierKD.

Siamese trackers are shown to be vulnerable to adversarial attacks recently. However, the existing attack methods craft the perturbations for each video independently, which comes at a non-negligible computational cost. In this paper, we show the existence of universal perturbations that can enable the targeted attack, e.g., forcing a tracker to follow the ground-truth trajectory with specified offsets, to be video-agnostic and free from inference in a network. Specifically, we attack a tracker by adding a universal translucent perturbation to the template image and adding afake target, i.e., a small universal adversarial patch, into the search images adhering to the predefined trajectory, so that the tracker outputs the location and size of thefake targetinstead of the real target. Our approach allows perturbing a novel video to come at no additional cost except the mere addition operations – and not require gradient optimization or network inference. Experimental results on several datasets demonstrate that our approach can effectively fool the Siamese trackers in a targeted attack manner. We show that the proposed perturbations are not only universal across videos, but also generalize well across different trackers. Such perturbations are therefore doubly universal, both with respect to the data and the network architectures. Our code is available athttps://github.com/lizhenbang56/SiamAttack.

Deep learning methods require massive of annotated data for optimizing parameters. For example, datasets attached with accurate bounding box annotations are essential for modern object detection tasks. However, labeling with such pixel-wise accuracy is laborious and time-consuming, and elaborate labeling procedures are indispensable for reducing man-made noise, involving annotation review and acceptance testing. In this paper, we focus on the impact of noisy location annotations on the performance of object detection approaches and aim to, on the user side, reduce the adverse effect of the noise. First, noticeable performance degradation is experimentally observed for both one-stage and two-stage detectors when noise is introduced to the bounding box annotations. For instance, our synthesized noise results in performance decrease from 38.9% AP to 33.6% AP for FCOS detector on COCO test split, and 37.8%AP to 33.7%AP for Faster R-CNN. Second, a self-correction technique based on a Bayesian filter for prediction ensemble is proposed to better exploit the noisy location annotations following a Teacher-Student learning paradigm. Experiments for both synthesized and real-world scenarios consistently demonstrate the effectiveness of our approach, e.g., our method increases the degraded performance of the FCOS detector from 33.6% AP to 35.6% AP on COCO.

Recent one-stage object detectors follow a per-pixel prediction approach that predicts both the object category scores and boundary positions from every single grid location. However, the most suitable positions for inferring different targets, i.e., the object category and boundaries, are generally different. Predicting all these targets from the same grid location thus may lead to sub-optimal results. In this paper, we analyze the suitable inference positions for object category and boundaries, and propose a prediction-target-decoupled detector named PDNet to establish a more flexible detection paradigm. Our PDNet with the prediction decoupling mechanism encodes different targets separately in different locations. A learnable prediction collection module is devised with two sets of dynamic points, i.e., dynamic boundary points and semantic points, to collect and aggregate the predictions from the favorable regions for localization and classification. We adopt a two-step strategy to learn these dynamic point positions, where the prior positions are estimated for different targets first, and the network further predicts residual offsets to the positions with better perceptions of the object properties. Extensive experiments on the MS COCO benchmark demonstrate the effectiveness and efficiency of our method. With a single ResNeXt-64x4d-101-DCN as the backbone, our detector achieves 50.1 AP with single-scale testing, which outperforms the state-of-the-art methods by an appreciable margin under the same experimental settings. Moreover, our detector is highly efficient as a one-stage framework. Our code will be public.

Object detection and instance segmentation are two fundamental computer vision tasks. They are closely correlated but their relations have not yet been fully explored in most previous work. This paper presents RDSNet, a novel deep architecture for reciprocal object detection and instance segmentation. To reciprocate these two tasks, we design a two-stream structure to learn features on both the object level (i.e., bounding boxes) and the pixel level (i.e., instance masks) jointly. Within this structure, information from the two streams is fused alternately, namely information on the object level introduces the awareness of instance and translation variance to the pixel level, and information on the pixel level refines the localization accuracy of objects on the object level in return. Specifically, a correlation module and a cropping module are proposed to yield instance masks, as well as a mask based boundary refinement module for more accurate bounding boxes. Extensive experimental analyses and comparisons on the COCO dataset demonstrate the effectiveness and efficiency of RDSNet. The source code is available at https://github.com/wangsr126/RDSNet.