Yuan Zhang

Tao Zhang, Xingye Tian, Yu Wu et al.

Video instance segmentation (VIS) is a critical task with diverse applications, including autonomous driving and video editing. Existing methods often underperform on complex and long videos in real world, primarily due to two factors. Firstly, offline methods are limited by the tightly-coupled modeling paradigm, which treats all frames equally and disregards the interdependencies between adjacent frames. Consequently, this leads to the introduction of excessive noise during long-term temporal alignment. Secondly, online methods suffer from inadequate utilization of temporal information. To tackle these challenges, we propose a decoupling strategy for VIS by dividing it into three independent sub-tasks: segmentation, tracking, and refinement. The efficacy of the decoupling strategy relies on two crucial elements: 1) attaining precise long-term alignment outcomes via frame-by-frame association during tracking, and 2) the effective utilization of temporal information predicated on the aforementioned accurate alignment outcomes during refinement. We introduce a novel referring tracker and temporal refiner to construct the \textbf{D}ecoupled \textbf{VIS} framework (\textbf{DVIS}). DVIS achieves new SOTA performance in both VIS and VPS, surpassing the current SOTA methods by 7.3 AP and 9.6 VPQ on the OVIS and VIPSeg datasets, which are the most challenging and realistic benchmarks. Moreover, thanks to the decoupling strategy, the referring tracker and temporal refiner are super light-weight (only 1.69\% of the segmenter FLOPs), allowing for efficient training and inference on a single GPU with 11G memory. The code is available at \href{https://github.com/zhang-tao-whu/DVIS}{https://github.com/zhang-tao-whu/DVIS}.

Tao Zhang, Xingye Tian, Haoran Wei et al.

Video panoptic segmentation is a challenging task that serves as the cornerstone of numerous downstream applications, including video editing and autonomous driving. We believe that the decoupling strategy proposed by DVIS enables more effective utilization of temporal information for both "thing" and "stuff" objects. In this report, we successfully validated the effectiveness of the decoupling strategy in video panoptic segmentation. Finally, our method achieved a VPQ score of 51.4 and 53.7 in the development and test phases, respectively, and ultimately ranked 1st in the VPS track of the 2nd PVUW Challenge. The code is available at https://github.com/zhang-tao-whu/DVIS

Tao Zhang, Xingye Tian, Yikang Zhou et al.

Video instance segmentation is a challenging task that serves as the cornerstone of numerous downstream applications, including video editing and autonomous driving. In this report, we present further improvements to the SOTA VIS method, DVIS. First, we introduce a denoising training strategy for the trainable tracker, allowing it to achieve more stable and accurate object tracking in complex and long videos. Additionally, we explore the role of visual foundation models in video instance segmentation. By utilizing a frozen VIT-L model pre-trained by DINO v2, DVIS demonstrates remarkable performance improvements. With these enhancements, our method achieves 57.9 AP and 56.0 AP in the development and test phases, respectively, and ultimately ranked 1st in the VIS track of the 5th LSVOS Challenge. The code will be available at https://github.com/zhang-tao-whu/DVIS.

Xianzhe Xu, Yiqi Jiang, Weihua Chen et al.

In this report, we present a fast and accurate object detection method dubbed DAMO-YOLO, which achieves higher performance than the state-of-the-art YOLO series. DAMO-YOLO is extended from YOLO with some new technologies, including Neural Architecture Search (NAS), efficient Reparameterized Generalized-FPN (RepGFPN), a lightweight head with AlignedOTA label assignment, and distillation enhancement. In particular, we use MAE-NAS, a method guided by the principle of maximum entropy, to search our detection backbone under the constraints of low latency and high performance, producing ResNet/CSP-like structures with spatial pyramid pooling and focus modules. In the design of necks and heads, we follow the rule of ``large neck, small head''.We import Generalized-FPN with accelerated queen-fusion to build the detector neck and upgrade its CSPNet with efficient layer aggregation networks (ELAN) and reparameterization. Then we investigate how detector head size affects detection performance and find that a heavy neck with only one task projection layer would yield better results.In addition, AlignedOTA is proposed to solve the misalignment problem in label assignment. And a distillation schema is introduced to improve performance to a higher level. Based on these new techs, we build a suite of models at various scales to meet the needs of different scenarios. For general industry requirements, we propose DAMO-YOLO-T/S/M/L. They can achieve 43.6/47.7/50.2/51.9 mAPs on COCO with the latency of 2.78/3.83/5.62/7.95 ms on T4 GPUs respectively. Additionally, for edge devices with limited computing power, we have also proposed DAMO-YOLO-Ns/Nm/Nl lightweight models. They can achieve 32.3/38.2/40.5 mAPs on COCO with the latency of 4.08/5.05/6.69 ms on X86-CPU. Our proposed general and lightweight models have outperformed other YOLO series models in their respective application scenarios.

Tao Zhang, Xingye Tian, Yikang Zhou et al.

We present the \textbf{D}ecoupled \textbf{VI}deo \textbf{S}egmentation (DVIS) framework, a novel approach for the challenging task of universal video segmentation, including video instance segmentation (VIS), video semantic segmentation (VSS), and video panoptic segmentation (VPS). Unlike previous methods that model video segmentation in an end-to-end manner, our approach decouples video segmentation into three cascaded sub-tasks: segmentation, tracking, and refinement. This decoupling design allows for simpler and more effective modeling of the spatio-temporal representations of objects, especially in complex scenes and long videos. Accordingly, we introduce two novel components: the referring tracker and the temporal refiner. These components track objects frame by frame and model spatio-temporal representations based on pre-aligned features. To improve the tracking capability of DVIS, we propose a denoising training strategy and introduce contrastive learning, resulting in a more robust framework named DVIS++. Furthermore, we evaluate DVIS++ in various settings, including open vocabulary and using a frozen pre-trained backbone. By integrating CLIP with DVIS++, we present OV-DVIS++, the first open-vocabulary universal video segmentation framework. We conduct extensive experiments on six mainstream benchmarks, including the VIS, VSS, and VPS datasets. Using a unified architecture, DVIS++ significantly outperforms state-of-the-art specialized methods on these benchmarks in both close- and open-vocabulary settings. Code:~\url{https://github.com/zhang-tao-whu/DVIS_Plus}.

Shaohua Pan, Xinyu Yi, Yan Zhou et al.

Combining sparse IMUs and a monocular camera is a new promising setting to perform real-time human motion capture. This paper proposes a diffusion-based solution to learn human motion priors and fuse the two modalities of signals together seamlessly in a unified framework. By delicately considering the characteristics of the two signals, the sequential visual information is considered as a whole and transformed into a condition embedding, while the inertial measurement is concatenated with the noisy body pose frame by frame to construct a sequential input for the diffusion model. Firstly, we observe that the visual information may be unavailable in some frames due to occlusions or subjects moving out of the camera view. Thus incorporating the sequential visual features as a whole to get a single feature embedding is robust to the occasional degenerations of visual information in those frames. On the other hand, the IMU measurements are robust to occlusions and always stable when signal transmission has no problem. So incorporating them frame-wisely could better explore the temporal information for the system. Experiments have demonstrated the effectiveness of the system design and its state-of-the-art performance in pose estimation compared with the previous works. Our codes are available for research at https://shaohua-pan.github.io/diffcap-page.

Tao Huang, Yuan Zhang, Mingkai Zheng et al.

The representation gap between teacher and student is an emerging topic in knowledge distillation (KD). To reduce the gap and improve the performance, current methods often resort to complicated training schemes, loss functions, and feature alignments, which are task-specific and feature-specific. In this paper, we state that the essence of these methods is to discard the noisy information and distill the valuable information in the feature, and propose a novel KD method dubbed DiffKD, to explicitly denoise and match features using diffusion models. Our approach is based on the observation that student features typically contain more noises than teacher features due to the smaller capacity of student model. To address this, we propose to denoise student features using a diffusion model trained by teacher features. This allows us to perform better distillation between the refined clean feature and teacher feature. Additionally, we introduce a light-weight diffusion model with a linear autoencoder to reduce the computation cost and an adaptive noise matching module to improve the denoising performance. Extensive experiments demonstrate that DiffKD is effective across various types of features and achieves state-of-the-art performance consistently on image classification, object detection, and semantic segmentation tasks. Code is available at https://github.com/hunto/DiffKD.

Yuan Zhang, Weihua Chen, Yichen Lu et al.

Knowledge distillation is an effective paradigm for boosting the performance of pocket-size model, especially when multiple teacher models are available, the student would break the upper limit again. However, it is not economical to train diverse teacher models for the disposable distillation. In this paper, we introduce a new concept dubbed Avatars for distillation, which are the inference ensemble models derived from the teacher. Concretely, (1) For each iteration of distillation training, various Avatars are generated by a perturbation transformation. We validate that Avatars own higher upper limit of working capacity and teaching ability, aiding the student model in learning diverse and receptive knowledge perspectives from the teacher model. (2) During the distillation, we propose an uncertainty-aware factor from the variance of statistical differences between the vanilla teacher and Avatars, to adjust Avatars' contribution on knowledge transfer adaptively. Avatar Knowledge Distillation AKD is fundamentally different from existing methods and refines with the innovative view of unequal training. Comprehensive experiments demonstrate the effectiveness of our Avatars mechanism, which polishes up the state-of-the-art distillation methods for dense prediction without more extra computational cost. The AKD brings at most 0.7 AP gains on COCO 2017 for Object Detection and 1.83 mIoU gains on Cityscapes for Semantic Segmentation, respectively. Code is available at https://github.com/Gumpest/AvatarKD.

Xingyu Chen, Yufeng Liu, Yajiao Dong et al.

In this work, we propose a framework for single-view hand mesh reconstruction, which can simultaneously achieve high reconstruction accuracy, fast inference speed, and temporal coherence. Specifically, for 2D encoding, we propose lightweight yet effective stacked structures. Regarding 3D decoding, we provide an efficient graph operator, namely depth-separable spiral convolution. Moreover, we present a novel feature lifting module for bridging the gap between 2D and 3D representations. This module begins with a map-based position regression (MapReg) block to integrate the merits of both heatmap encoding and position regression paradigms for improved 2D accuracy and temporal coherence. Furthermore, MapReg is followed by pose pooling and pose-to-vertex lifting approaches, which transform 2D pose encodings to semantic features of 3D vertices. Overall, our hand reconstruction framework, called MobRecon, comprises affordable computational costs and miniature model size, which reaches a high inference speed of 83FPS on Apple A14 CPU. Extensive experiments on popular datasets such as FreiHAND, RHD, and HO3Dv2 demonstrate that our MobRecon achieves superior performance on reconstruction accuracy and temporal coherence. Our code is publicly available at https://github.com/SeanChenxy/HandMesh.

Zhi Qiao, Yu Zhou, Jin Wei et al.

Nowadays, scene text recognition has attracted more and more attention due to its various applications. Most state-of-the-art methods adopt an encoder-decoder framework with attention mechanism, which generates text autoregressively from left to right. Despite the convincing performance, the speed is limited because of the one-by-one decoding strategy. As opposed to autoregressive models, non-autoregressive models predict the results in parallel with a much shorter inference time, but the accuracy falls behind the autoregressive counterpart considerably. In this paper, we propose a Parallel, Iterative and Mimicking Network (PIMNet) to balance accuracy and efficiency. Specifically, PIMNet adopts a parallel attention mechanism to predict the text faster and an iterative generation mechanism to make the predictions more accurate. In each iteration, the context information is fully explored. To improve learning of the hidden layer, we exploit the mimicking learning in the training phase, where an additional autoregressive decoder is adopted and the parallel decoder mimics the autoregressive decoder with fitting outputs of the hidden layer. With the shared backbone between the two decoders, the proposed PIMNet can be trained end-to-end without pre-training. During inference, the branch of the autoregressive decoder is removed for a faster speed. Extensive experiments on public benchmarks demonstrate the effectiveness and efficiency of PIMNet. Our code will be available at https://github.com/Pay20Y/PIMNet.

Defang Chen, Jian-Ping Mei, Yuan Zhang et al.

Knowledge distillation is a technique to enhance the generalization ability of a student model by exploiting outputs from a teacher model. Recently, feature-map based variants explore knowledge transfer between manually assigned teacher-student pairs in intermediate layers for further improvement. However, layer semantics may vary in different neural networks and semantic mismatch in manual layer associations will lead to performance degeneration due to negative regularization. To address this issue, we propose Semantic Calibration for cross-layer Knowledge Distillation (SemCKD), which automatically assigns proper target layers of the teacher model for each student layer with an attention mechanism. With a learned attention distribution, each student layer distills knowledge contained in multiple teacher layers rather than a specific intermediate layer for appropriate cross-layer supervision. We further provide theoretical analysis of the association weights and conduct extensive experiments to demonstrate the effectiveness of our approach. Code is avaliable at \url{https://github.com/DefangChen/SemCKD}.

Haoxiang Tian, Xingshuo Han, Guoquan Wu et al.

Multi-objective evolutionary algorithms (MOEAs) are widely used for searching optimal solutions in complex multi-component applications. Traditional MOEAs for multi-component deep learning (MCDL) systems face challenges in enhancing the search efficiency while maintaining the diversity. To combat these, this paper proposes $μ$MOEA, the first LLM-empowered adaptive evolutionary search algorithm to detect safety violations in MCDL systems. Inspired by the context-understanding ability of Large Language Models (LLMs), $μ$MOEA promotes the LLM to comprehend the optimization problem and generate an initial population tailed to evolutionary objectives. Subsequently, it employs adaptive selection and variation to iteratively produce offspring, balancing the evolutionary efficiency and diversity. During the evolutionary process, to navigate away from the local optima, $μ$MOEA integrates the evolutionary experience back into the LLM. This utilization harnesses the LLM's quantitative reasoning prowess to generate differential seeds, breaking away from current optimal solutions. We evaluate $μ$MOEA in finding safety violations of MCDL systems, and compare its performance with state-of-the-art MOEA methods. Experimental results show that $μ$MOEA can significantly improve the efficiency and diversity of the evolutionary search.

Yuan Zhang, Jian Cao, Ling Zhang et al.

Learning subtle representation about object parts plays a vital role in fine-grained visual recognition (FGVR) field. The vision transformer (ViT) achieves promising results on computer vision due to its attention mechanism. Nonetheless, with the fixed size of patches in ViT, the class token in deep layer focuses on the global receptive field and cannot generate multi-granularity features for FGVR. To capture region attention without box annotations and compensate for ViT shortcomings in FGVR, we propose a novel method named Adaptive attention multi-scale Fusion Transformer (AFTrans). The Selective Attention Collection Module (SACM) in our approach leverages attention weights in ViT and filters them adaptively to correspond with the relative importance of input patches. The multiple scales (global and local) pipeline is supervised by our weights sharing encoder and can be easily trained end-to-end. Comprehensive experiments demonstrate that AFTrans can achieve SOTA performance on three published fine-grained benchmarks: CUB-200-2011, Stanford Dogs and iNat2017.

Xiangcheng Liu, Jian Cao, Hongyi Yao et al.

Channel pruning is one of the major compression approaches for deep neural networks. While previous pruning methods have mostly focused on identifying unimportant channels, channel pruning is considered as a special case of neural architecture search in recent years. However, existing methods are either complicated or prone to sub-optimal pruning. In this paper, we propose a pruning framework that adaptively determines the number of each layer's channels as well as the wights inheritance criteria for sub-network. Firstly, evaluate the importance of each block in the network based on the mean of the scaling parameters of the BN layers. Secondly, use the bisection method to quickly find the compact sub-network satisfying the budget. Finally, adaptively and efficiently choose the weight inheritance criterion that fits the current architecture and fine-tune the pruned network to recover performance. AdaPruner allows to obtain pruned network quickly, accurately and efficiently, taking into account both the structure and initialization weights. We prune the currently popular CNN models (VGG, ResNet, MobileNetV2) on different image classification datasets, and the experimental results demonstrate the effectiveness of our proposed method. On ImageNet, we reduce 32.8% FLOPs of MobileNetV2 with only 0.62% decrease for top-1 accuracy, which exceeds all previous state-of-the-art channel pruning methods. The code will be released.

Wen Gao, Shan Liu, Xiaozhong Xu et al.

In recent years, video data has dominated internet traffic and becomes one of the major data formats. With the emerging 5G and internet of things (IoT) technologies, more and more videos are generated by edge devices, sent across networks, and consumed by machines. The volume of video consumed by machine is exceeding the volume of video consumed by humans. Machine vision tasks include object detection, segmentation, tracking, and other machine-based applications, which are quite different from those for human consumption. On the other hand, due to large volumes of video data, it is essential to compress video before transmission. Thus, efficient video coding for machines (VCM) has become an important topic in academia and industry. In July 2019, the international standardization organization, i.e., MPEG, created an Ad-Hoc group named VCM to study the requirements for potential standardization work. In this paper, we will address the recent development activities in the MPEG VCM group. Specifically, we will first provide an overview of the MPEG VCM group including use cases, requirements, processing pipelines, plan for potential VCM standards, followed by the evaluation framework including machine-vision tasks, dataset, evaluation metrics, and anchor generation. We then introduce technology solutions proposed so far and discuss the recent responses to the Call for Evidence issued by MPEG VCM group.

Yang Li, Jiacong He, Xin Zhou et al.

We present a new problem: grounding natural language instructions to mobile user interface actions, and create three new datasets for it. For full task evaluation, we create PIXELHELP, a corpus that pairs English instructions with actions performed by people on a mobile UI emulator. To scale training, we decouple the language and action data by (a) annotating action phrase spans in HowTo instructions and (b) synthesizing grounded descriptions of actions for mobile user interfaces. We use a Transformer to extract action phrase tuples from long-range natural language instructions. A grounding Transformer then contextually represents UI objects using both their content and screen position and connects them to object descriptions. Given a starting screen and instruction, our model achieves 70.59% accuracy on predicting complete ground-truth action sequences in PIXELHELP.

Rang Meng, Weijie Chen, Di Xie et al.

Layer assignment is seldom picked out as an independent research topic in neural architecture search. In this paper, for the first time, we systematically investigate the impact of different layer assignments to the network performance by building an architecture dataset of layer assignment on CIFAR-100. Through analyzing this dataset, we discover a neural inheritance relation among the networks with different layer assignments, that is, the optimal layer assignments for deeper networks always inherit from those for shallow networks. Inspired by this neural inheritance relation, we propose an efficient one-shot layer assignment search approach via inherited sampling. Specifically, the optimal layer assignment searched in the shallow network can be provided as a strong sampling priori to train and search the deeper ones in supernet, which extremely reduces the network search space. Comprehensive experiments carried out on CIFAR-100 illustrate the efficiency of our proposed method. Our search results are strongly consistent with the optimal ones directly selected from the architecture dataset. To further confirm the generalization of our proposed method, we also conduct experiments on Tiny-ImageNet and ImageNet. Our searched results are remarkably superior to the handcrafted ones under the unchanged computational budgets. The neural inheritance relation discovered in this paper can provide insights to the universal neural architecture search.

Gabriel Ilharco, Yuan Zhang, Jason Baldridge

Systems that can associate images with their spoken audio captions are an important step towards visually grounded language learning. We describe a scalable method to automatically generate diverse audio for image captioning datasets. This supports pretraining deep networks for encoding both audio and images, which we do via a dual encoder that learns to align latent representations from both modalities. We show that a masked margin softmax loss for such models is superior to the standard triplet loss. We fine-tune these models on the Flickr8k Audio Captions Corpus and obtain state-of-the-art results---improving recall in the top 10 from 29.6% to 49.5%. We also obtain human ratings on retrieval outputs to better assess the impact of incidentally matching image-caption pairs that were not associated in the data, finding that automatic evaluation substantially underestimates the quality of the retrieved results.

Weijie Chen, Di Xie, Yuan Zhang et al.

Shift operation is an efficient alternative over depthwise separable convolution. However, it is still bottlenecked by its implementation manner, namely memory movement. To put this direction forward, a new and novel basic component named Sparse Shift Layer (SSL) is introduced in this paper to construct efficient convolutional neural networks. In this family of architectures, the basic block is only composed by 1x1 convolutional layers with only a few shift operations applied to the intermediate feature maps. To make this idea feasible, we introduce shift operation penalty during optimization and further propose a quantization-aware shift learning method to impose the learned displacement more friendly for inference. Extensive ablation studies indicate that only a few shift operations are sufficient to provide spatial information communication. Furthermore, to maximize the role of SSL, we redesign an improved network architecture to Fully Exploit the limited capacity of neural Network (FE-Net). Equipped with SSL, this network can achieve 75.0% top-1 accuracy on ImageNet with only 563M M-Adds. It surpasses other counterparts constructed by depthwise separable convolution and the networks searched by NAS in terms of accuracy and practical speed.

Weijie Chen, Yuan Zhang, Di Xie et al.

Neuron pruning is an efficient method to compress the network into a slimmer one for reducing the computational cost and storage overhead. Most of state-of-the-art results are obtained in a layer-by-layer optimization mode. It discards the unimportant input neurons and uses the survived ones to reconstruct the output neurons approaching to the original ones in a layer-by-layer manner. However, an unnoticed problem arises that the information loss is accumulated as layer increases since the survived neurons still do not encode the entire information as before. A better alternative is to propagate the entire useful information to reconstruct the pruned layer instead of directly discarding the less important neurons. To this end, we propose a novel Layer Decomposition-Recomposition Framework (LDRF) for neuron pruning, by which each layer's output information is recovered in an embedding space and then propagated to reconstruct the following pruned layers with useful information preserved. We mainly conduct our experiments on ILSVRC-12 benchmark with VGG-16 and ResNet-50. What should be emphasized is that our results before end-to-end fine-tuning are significantly superior owing to the information-preserving property of our proposed framework.With end-to-end fine-tuning, we achieve state-of-the-art results of 5.13x and 3x speed-up with only 0.5% and 0.65% top-5 accuracy drop respectively, which outperform the existing neuron pruning methods.