Zhenyu Wu

Zhenyu Wu, Lin Wang, Wei Wang et al.

Although deep salient object detection (SOD) has achieved remarkable progress, deep SOD models are extremely data-hungry, requiring large-scale pixel-wise annotations to deliver such promising results. In this paper, we propose a novel yet effective method for SOD, coined SODGAN, which can generate infinite high-quality image-mask pairs requiring only a few labeled data, and these synthesized pairs can replace the human-labeled DUTS-TR to train any off-the-shelf SOD model. Its contribution is three-fold. 1) Our proposed diffusion embedding network can address the manifold mismatch and is tractable for the latent code generation, better matching with the ImageNet latent space. 2) For the first time, our proposed few-shot saliency mask generator can synthesize infinite accurate image synchronized saliency masks with a few labeled data. 3) Our proposed quality-aware discriminator can select highquality synthesized image-mask pairs from noisy synthetic data pool, improving the quality of synthetic data. For the first time, our SODGAN tackles SOD with synthetic data directly generated from the generative model, which opens up a new research paradigm for SOD. Extensive experimental results show that the saliency model trained on synthetic data can achieve $98.4\%$ F-measure of the saliency model trained on the DUTS-TR. Moreover, our approach achieves a new SOTA performance in semi/weakly-supervised methods, and even outperforms several fully-supervised SOTA methods. Code is available at https://github.com/wuzhenyubuaa/SODGAN

Zhenyu Hu, Zhenyu Wu, Pengcheng Pi et al.

Unmanned Aerial Vehicles (UAVs) based video text spotting has been extensively used in civil and military domains. UAV's limited battery capacity motivates us to develop an energy-efficient video text spotting solution. In this paper, we first revisit RCNN's crop & resize training strategy and empirically find that it outperforms aligned RoI sampling on a real-world video text dataset captured by UAV. To reduce energy consumption, we further propose a multi-stage image processor that takes videos' redundancy, continuity, and mixed degradation into account. Lastly, the model is pruned and quantized before deployed on Raspberry Pi. Our proposed energy-efficient video text spotting solution, dubbed as E^2VTS, outperforms all previous methods by achieving a competitive tradeoff between energy efficiency and performance. All our codes and pre-trained models are available at https://github.com/wuzhenyusjtu/LPCVC20-VideoTextSpotting.

Zhenyu Wu, Shuai Li, Chenglizhao Chen et al.

Recent research advances in salient object detection (SOD) could largely be attributed to ever-stronger multi-scale feature representation empowered by the deep learning technologies. The existing SOD deep models extract multi-scale features via the off-the-shelf encoders and combine them smartly via various delicate decoders. However, the kernel sizes in this commonly-used thread are usually "fixed". In our new experiments, we have observed that kernels of small size are preferable in scenarios containing tiny salient objects. In contrast, large kernel sizes could perform better for images with large salient objects. Inspired by this observation, we advocate the "dynamic" scale routing (as a brand-new idea) in this paper. It will result in a generic plug-in that could directly fit the existing feature backbone. This paper's key technical innovations are two-fold. First, instead of using the vanilla convolution with fixed kernel sizes for the encoder design, we propose the dynamic pyramid convolution (DPConv), which dynamically selects the best-suited kernel sizes w.r.t. the given input. Second, we provide a self-adaptive bidirectional decoder design to accommodate the DPConv-based encoder best. The most significant highlight is its capability of routing between feature scales and their dynamic collection, making the inference process scale-aware. As a result, this paper continues to enhance the current SOTA performance. Both the code and dataset are publicly available at https://github.com/wuzhenyubuaa/DPNet.

Zhenyu Wu, Ziwei Wang, Xiuwei Xu et al.

Equipping embodied agents with commonsense is important for robots to successfully complete complex human instructions in general environments. Recent large language models (LLM) can embed rich semantic knowledge for agents in plan generation of complex tasks, while they lack the information about the realistic world and usually yield infeasible action sequences. In this paper, we propose a TAsk Planing Agent (TaPA) in embodied tasks for grounded planning with physical scene constraint, where the agent generates executable plans according to the existed objects in the scene by aligning LLMs with the visual perception models. Specifically, we first construct a multimodal dataset containing triplets of indoor scenes, instructions and action plans, where we provide the designed prompts and the list of existing objects in the scene for GPT-3.5 to generate a large number of instructions and corresponding planned actions. The generated data is leveraged for grounded plan tuning of pre-trained LLMs. During inference, we discover the objects in the scene by extending open-vocabulary object detectors to multi-view RGB images collected in different achievable locations. Experimental results show that the generated plan from our TaPA framework can achieve higher success rate than LLaVA and GPT-3.5 by a sizable margin, which indicates the practicality of embodied task planning in general and complex environments.

Priya Narayanan, Xin Hu, Zhenyu Wu et al.

Imagery collected from outdoor visual environments is often degraded due to the presence of dense smoke or haze. A key challenge for research in scene understanding in these degraded visual environments (DVE) is the lack of representative benchmark datasets. These datasets are required to evaluate state-of-the-art vision algorithms (e.g., detection and tracking) in degraded settings. In this paper, we address some of these limitations by introducing the first realistic hazy image benchmark, from both aerial and ground view, with paired haze-free images, and in-situ haze density measurements. This dataset was produced in a controlled environment with professional smoke generating machines that covered the entire scene, and consists of images captured from the perspective of both an unmanned aerial vehicle (UAV) and an unmanned ground vehicle (UGV). We also evaluate a set of representative state-of-the-art dehazing approaches as well as object detectors on the dataset. The full dataset presented in this paper, including the ground truth object classification bounding boxes and haze density measurements, is provided for the community to evaluate their algorithms at: https://a2i2-archangel.vision. A subset of this dataset has been used for the ``Object Detection in Haze'' Track of CVPR UG2 2022 challenge at http://cvpr2022.ug2challenge.org/track1.html.

Xin Hu, Zhenyu Wu, Hao-Yu Miao et al.

Video action detection (spatio-temporal action localization) is usually the starting point for human-centric intelligent analysis of videos nowadays. It has high practical impacts for many applications across robotics, security, healthcare, etc. The two-stage paradigm of Faster R-CNN inspires a standard paradigm of video action detection in object detection, i.e., firstly generating person proposals and then classifying their actions. However, none of the existing solutions could provide fine-grained action detection to the "who-when-where-what" level. This paper presents a tracking-based solution to accurately and efficiently localize predefined key actions spatially (by predicting the associated target IDs and locations) and temporally (by predicting the time in exact frame indices). This solution won first place in the UAV-Video Track of 2021 Low-Power Computer Vision Challenge (LPCVC).

Runhao Li, Wenkai Guo, Zhenyu Wu et al.

Pre-trained Vision-Language-Action (VLA) models have achieved remarkable success in improving robustness and generalization for end-to-end robotic manipulation. However, these models struggle with long-horizon tasks due to their lack of memory and reliance solely on immediate sensory inputs. To address this limitation, we propose Memory-Augmented Prompting for Vision-Language-Action model (MAP-VLA), a novel framework that empowers pre-trained VLA models with demonstration-derived memory prompts to augment action generation for long-horizon robotic manipulation tasks. To achieve this, MAP-VLA first constructs a memory library from historical demonstrations, where each memory unit captures information about a specific stage of a task. These memory units are implemented as learnable soft prompts optimized through prompt tuning. Then, during real-time task execution, MAP-VLA retrieves relevant memory through trajectory similarity matching and dynamically integrates it into the VLA model for augmented action generation. Importantly, this prompt tuning and retrieval augmentation approach operates as a plug-and-play module for a frozen VLA model, offering a lightweight and flexible solution to improve task performance. Experimental results show that MAP-VLA delivers up to 7.0% absolute performance gains in the simulation benchmark and 25.0% on real robot evaluations for long-horizon tasks, surpassing the current state-of-the-art methods.

Kumara Kahatapitiya, Zhou Ren, Haoxiang Li et al.

Temporal Activity Detection aims to predict activity classes per frame, in contrast to video-level predictions in Activity Classification (i.e., Activity Recognition). Due to the expensive frame-level annotations required for detection, the scale of detection datasets is limited. Thus, commonly, previous work on temporal activity detection resorts to fine-tuning a classification model pretrained on large-scale classification datasets (e.g., Kinetics-400). However, such pretrained models are not ideal for downstream detection, due to the disparity between the pretraining and the downstream fine-tuning tasks. In this work, we propose a novel 'weakly-guided self-supervised' pretraining method for detection. We leverage weak labels (classification) to introduce a self-supervised pretext task (detection) by generating frame-level pseudo labels, multi-action frames, and action segments. Simply put, we design a detection task similar to downstream, on large-scale classification data, without extra annotations. We show that the models pretrained with the proposed weakly-guided self-supervised detection task outperform prior work on multiple challenging activity detection benchmarks, including Charades and MultiTHUMOS. Our extensive ablations further provide insights on when and how to use the proposed models for activity detection. Code is available at https://github.com/kkahatapitiya/SSDet.

Zhenyu Wu, Zhaowen Wang, Ye Yuan et al.

Generative adversarial networks (GANs) nowadays are capable of producing images of incredible realism. One concern raised is whether the state-of-the-art GAN's learned distribution still suffers from mode collapse, and what to do if so. Existing diversity tests of samples from GANs are usually conducted qualitatively on a small scale, and/or depends on the access to original training data as well as the trained model parameters. This paper explores to diagnose GAN intra-mode collapse and calibrate that, in a novel black-box setting: no access to training data, nor the trained model parameters, is assumed. The new setting is practically demanded, yet rarely explored and significantly more challenging. As a first stab, we devise a set of statistical tools based on sampling, that can visualize, quantify, and rectify intra-mode collapse. We demonstrate the effectiveness of our proposed diagnosis and calibration techniques, via extensive simulations and experiments, on unconditional GAN image generation (e.g., face and vehicle). Our study reveals that the intra-mode collapse is still a prevailing problem in state-of-the-art GANs and the mode collapse is diagnosable and calibratable in black-box settings. Our codes are available at: https://github.com/VITA-Group/BlackBoxGANCollapse.

Yingying Jiao, Zhigang Wang, Zhenguang Liu et al.

Human pose estimation has given rise to a broad spectrum of novel and compelling applications, including action recognition, sports analysis, as well as surveillance. However, accurate video pose estimation remains an open challenge. One aspect that has been overlooked so far is that existing methods learn motion clues from all pixels rather than focusing on the target human body, making them easily misled and disrupted by unimportant information such as background changes or movements of other people. Additionally, while the current Transformer-based pose estimation methods has demonstrated impressive performance with global modeling, they struggle with local context perception and precise positional identification. In this paper, we try to tackle these challenges from three aspects: (1) We propose a bilayer Human-Keypoint Mask module that performs coarse-to-fine visual token refinement, which gradually zooms in on the target human body and keypoints while masking out unimportant figure regions. (2) We further introduce a novel deformable cross attention mechanism and a bidirectional separation strategy to adaptively aggregate spatial and temporal motion clues from constrained surrounding contexts. (3) We mathematically formulate the deformable cross attention, constraining that the model focuses solely on the regions centered at the target person body. Empirically, our method achieves state-of-the-art performance on three large-scale benchmark datasets. A remarkable highlight is that our method achieves an 84.8 mean Average Precision (mAP) on the challenging wrist joint, which significantly outperforms the 81.5 mAP achieved by the current state-of-the-art method on the PoseTrack2017 dataset.

Yingying Jiao, Zhigang Wang, Sifan Wu et al.

Human pose estimation in videos remains a challenge, largely due to the reliance on extensive manual annotation of large datasets, which is expensive and labor-intensive. Furthermore, existing approaches often struggle to capture long-range temporal dependencies and overlook the complementary relationship between temporal pose heatmaps and visual features. To address these limitations, we introduce STDPose, a novel framework that enhances human pose estimation by learning spatiotemporal dynamics in sparsely-labeled videos. STDPose incorporates two key innovations: 1) A novel Dynamic-Aware Mask to capture long-range motion context, allowing for a nuanced understanding of pose changes. 2) A system for encoding and aggregating spatiotemporal representations and motion dynamics to effectively model spatiotemporal relationships, improving the accuracy and robustness of pose estimation. STDPose establishes a new performance benchmark for both video pose propagation (i.e., propagating pose annotations from labeled frames to unlabeled frames) and pose estimation tasks, across three large-scale evaluation datasets. Additionally, utilizing pseudo-labels generated by pose propagation, STDPose achieves competitive performance with only 26.7% labeled data.

Liqiang Song, Ye Bi, Mengqiu Yao et al.

Social connections play a vital role in improving the performance of recommendation systems (RS). However, incorporating social information into RS is challenging. Most existing models usually consider social influences in a given session, ignoring that both users preferences and their friends influences are evolving. Moreover, in real world, social relations are sparse. Modeling dynamic influences and alleviating data sparsity is of great importance. In this paper, we propose a unified framework named Dynamic RElation Aware Model (DREAM) for social recommendation, which tries to model both users dynamic interests and their friends temporal influences. Specifically, we design temporal information encoding modules, because of which user representations are updated in each session. The updated user representations are transferred to relational-GAT modules, subsequently influence the operations on social networks. In each session, to solve social relation sparsity, we utilize glove-based method to complete social network with virtual friends. Then we employ relational-GAT module over completed social networks to update users representations. In the extensive experiments on the public datasets, DREAM significantly outperforms the state-of-the-art solutions.

Zhenyu Wu, Shuai Li, Chenglizhao Chen et al.

The real human attention is an interactive activity between our visual system and our brain, using both low-level visual stimulus and high-level semantic information. Previous image salient object detection (SOD) works conduct their saliency predictions in a multi-task manner, i.e., performing pixel-wise saliency regression and segmentation-like saliency refinement at the same time, which degenerates their feature backbones in revealing semantic information. However, given an image, we tend to pay more attention to those regions which are semantically salient even in the case that these regions are perceptually not the most salient ones at first glance. In this paper, we divide the SOD problem into two sequential tasks: 1) we propose a lightweight, weakly supervised deep network to coarsely locate those semantically salient regions first; 2) then, as a post-processing procedure, we selectively fuse multiple off-the-shelf deep models on these semantically salient regions as the pixel-wise saliency refinement. In sharp contrast to the state-of-the-art (SOTA) methods that focus on learning pixel-wise saliency in "single image" using perceptual clues mainly, our method has investigated the "object-level semantic ranks between multiple images", of which the methodology is more consistent with the real human attention mechanism. Our method is simple yet effective, which is the first attempt to consider the salient object detection mainly as an object-level semantic re-ranking problem.

Zhenyu Wu, Shuai Li, Chenglizhao Chen et al.

Fully convolutional networks have shown outstanding performance in the salient object detection (SOD) field. The state-of-the-art (SOTA) methods have a tendency to become deeper and more complex, which easily homogenize their learned deep features, resulting in a clear performance bottleneck. In sharp contrast to the conventional ``deeper'' schemes, this paper proposes a ``wider'' network architecture which consists of parallel sub networks with totally different network architectures. In this way, those deep features obtained via these two sub networks will exhibit large diversity, which will have large potential to be able to complement with each other. However, a large diversity may easily lead to the feature conflictions, thus we use the dense short-connections to enable a recursively interaction between the parallel sub networks, pursuing an optimal complementary status between multi-model deep features. Finally, all these complementary multi-model deep features will be selectively fused to make high-performance salient object detections. Extensive experiments on several famous benchmarks clearly demonstrate the superior performance, good generalization, and powerful learning ability of the proposed wider framework.

Zhenyu Wu, Shuai Li, Chenglizhao Chen et al.

Compared with the conventional hand-crafted approaches, the deep learning based methods have achieved tremendous performance improvements by training exquisitely crafted fancy networks over large-scale training sets. However, do we really need large-scale training set for salient object detection (SOD)? In this paper, we provide a deeper insight into the interrelationship between the SOD performances and the training sets. To alleviate the conventional demands for large-scale training data, we provide a feasible way to construct a novel small-scale training set, which only contains 4K images. Moreover, we propose a novel bi-stream network to take full advantage of our proposed small training set, which is consisted of two feature backbones with different structures, achieving complementary semantical saliency fusion via the proposed gate control unit. To our best knowledge, this is the first attempt to use a small-scale training set to outperform state-of-the-art models which are trained on large-scale training sets; nevertheless, our method can still achieve the leading state-of-the-art performance on five benchmark datasets.

Bo Huang, Ye Bi, Zhenyu Wu et al.

The problem of session-based recommendation aims to predict user next actions based on session histories. Previous methods models session histories into sequences and estimate user latent features by RNN and GNN methods to make recommendations. However under massive-scale and complicated financial recommendation scenarios with both virtual and real commodities , such methods are not sufficient to represent accurate user latent features and neglect the long-term characteristics of users. To take long-term preference and dynamic interests into account, we propose a novel method, i.e. User-Based Embeddings Recommendation with Graph Neural Network, UBER-GNN for brevity. UBER-GNN takes advantage of structured data to generate longterm user preferences, and transfers session sequences into graphs to generate graph-based dynamic interests. The final user latent feature is then represented as the composition of the long-term preferences and the dynamic interests using attention mechanism. Extensive experiments conducted on real Ping An scenario show that UBER-GNN outperforms the state-of-the-art session-based recommendation methods.