1 code implementation • 21 Mar 2024 • Han Zhao, Min Zhang, Wei Zhao, Pengxiang Ding, Siteng Huang, Donglin Wang
In recent years, the application of multimodal large language models (MLLM) in various fields has achieved remarkable success.
1 code implementation • 19 Jul 2023 • Yachen Kang, Li He, Jinxin Liu, Zifeng Zhuang, Donglin Wang
Due to the existence of similarity trap, such consistency regularization improperly enhances the consistency possiblity of the model's predictions between segment pairs, and thus reduces the confidence in reward learning, since the augmented distribution does not match with the original one in PbRL.
2 code implementations • 22 Feb 2023 • Zifeng Zhuang, Kun Lei, Jinxin Liu, Donglin Wang, Yilang Guo
Offline reinforcement learning (RL) is a challenging setting where existing off-policy actor-critic methods perform poorly due to the overestimation of out-of-distribution state-action pairs.
1 code implementation • CVPR 2023 • Siteng Huang, Biao Gong, Yulin Pan, Jianwen Jiang, Yiliang Lv, Yuyuan Li, Donglin Wang
Many recent studies leverage the pre-trained CLIP for text-video cross-modal retrieval by tuning the backbone with additional heavy modules, which not only brings huge computational burdens with much more parameters, but also leads to the knowledge forgetting from upstream models.
1 code implementation • 13 Aug 2023 • Yu Song, Donglin Wang
Graph Neural Networks (GNNs) are state-of-the-art models for performing prediction tasks on graphs.
1 code implementation • 25 May 2023 • Yachen Kang, Diyuan Shi, Jinxin Liu, Li He, Donglin Wang
Instead, the agent is provided with fixed offline trajectories and human preferences between pairs of trajectories to extract the dynamics and task information, respectively.
1 code implementation • 22 Jun 2023 • Jinxin Liu, Ziqi Zhang, Zhenyu Wei, Zifeng Zhuang, Yachen Kang, Sibo Gai, Donglin Wang
Offline reinforcement learning (RL) aims to learn a policy using only pre-collected and fixed data.
1 code implementation • 15 Dec 2023 • Shuanghao Bai, Min Zhang, Wanqi Zhou, Siteng Huang, Zhirong Luan, Donglin Wang, Badong Chen
Therefore, in this paper, we first experimentally demonstrate that the unsupervised-trained VLMs can significantly reduce the distribution discrepancy between source and target domains, thereby improving the performance of UDA.
1 code implementation • 19 Dec 2023 • Pengxiang Ding, Qiongjie Cui, Min Zhang, Mengyuan Liu, Haofan Wang, Donglin Wang
Human motion forecasting, with the goal of estimating future human behavior over a period of time, is a fundamental task in many real-world applications.
1 code implementation • 10 Sep 2020 • Siteng Huang, Min Zhang, Yachen Kang, Donglin Wang
However, these approaches only augment the representations of samples with available semantics while ignoring the query set, which loses the potential for the improvement and may lead to a shift between the modalities combination and the pure-visual representation.
1 code implementation • 14 Jul 2022 • Min Zhang, Siteng Huang, Wenbin Li, Donglin Wang
To solve this problem, we present a plug-in Hierarchical Tree Structure-aware (HTS) method, which not only learns the relationship of FSL and pretext tasks, but more importantly, can adaptively select and aggregate feature representations generated by pretext tasks to maximize the performance of FSL tasks.
1 code implementation • 27 Mar 2023 • Siteng Huang, Biao Gong, Yutong Feng, Min Zhang, Yiliang Lv, Donglin Wang
Recent compositional zero-shot learning (CZSL) methods adapt pre-trained vision-language models (VLMs) by constructing trainable prompts only for composed state-object pairs.
1 code implementation • 7 Jun 2021 • Xin Yang, Ning Zhang, Donglin Wang
Fourth, we generate three corresponding masks based on the 20 selected ROIs from group ICA, the 20 ROIs selected from dictionary learning, and the 40 combined ROIs selected from both.
1 code implementation • 3 Sep 2023 • Xuyang Liu, Siteng Huang, Yachen Kang, Honggang Chen, Donglin Wang
Large-scale text-to-image diffusion models have shown impressive capabilities for generative tasks by leveraging strong vision-language alignment from pre-training.
1 code implementation • 1 Oct 2023 • Teng Xiao, Zhengyu Chen, Donglin Wang, Suhang Wang
To compensate for this, in this paper, we present learning to propagate, a general learning framework that not only learns the GNN parameters for prediction but more importantly, can explicitly learn the interpretable and personalized propagate strategies for different nodes and various types of graphs.
1 code implementation • 15 Dec 2023 • Shuanghao Bai, Wanqi Zhou, Zhirong Luan, Donglin Wang, Badong Chen
Multilayer perceptron (MLP) has shown its capability to learn transferable representations in various downstream tasks, such as unsupervised image classification and supervised concept generalization.
1 code implementation • 13 Nov 2021 • Xintao Xiang, Tiancheng Huang, Donglin Wang
In this paper, we propose Learning to Evolve on Dynamic Graphs (LEDG) - a novel algorithm that jointly learns graph information and time information.
no code implementations • 10 Dec 2020 • Tiancheng Huang, Ke Xu, Donglin Wang
Domain adaptation using graph-structured networks learns label-discriminative and network-invariant node embeddings by sharing graph parameters.
no code implementations • 10 Dec 2020 • Ting Wang, Zongkai Wu, Donglin Wang
In the training phase, we first locate the generalization problem to the visual perception module, and then compare two meta-learning algorithms for better generalization in seen and unseen environments.
no code implementations • 11 Apr 2021 • Jinxin Liu, Donglin Wang, Qiangxing Tian, Zhengyu Chen
It is of significance for an agent to learn a widely applicable and general-purpose policy that can achieve diverse goals including images and text descriptions.
no code implementations • CVPR 2021 • Zhengyu Chen, Jixie Ge, Heshen Zhan, Siteng Huang, Donglin Wang
While few-shot learning (FSL) aims for rapid generalization to new concepts with little supervision, self-supervised learning (SSL) constructs supervisory signals directly computed from unlabeled data.
no code implementations • 27 Apr 2021 • Shiqi Chen, Zhengyu Chen, Donglin Wang
Meta Reinforcement Learning (MRL) enables an agent to learn from a limited number of past trajectories and extrapolate to a new task.
no code implementations • 29 Sep 2021 • Siteng Huang, Qiyao Wei, Donglin Wang
To narrow the considerable gap between artificial and human intelligence, we propose a new task, namely reference-limited compositional learning (RLCL), which reproduces three core challenges to mimic human perception: compositional learning, few-shot, and few referential compositions.
no code implementations • 15 Oct 2021 • Ryan Jacobs, Mingren Shen, YuHan Liu, Wei Hao, Xiaoshan Li, Ruoyu He, Jacob RC Greaves, Donglin Wang, Zeming Xie, Zitong Huang, Chao Wang, Kevin G. Field, Dane Morgan
In this work, we perform semantic segmentation of multiple defect types in electron microscopy images of irradiated FeCrAl alloys using a deep learning Mask Regional Convolutional Neural Network (Mask R-CNN) model.
no code implementations • 21 Oct 2021 • Yachen Kang, Jinxin Liu, Xin Cao, Donglin Wang
To achieve this, the widely used GAN-inspired IRL method is adopted, and its discriminator, recognizing policy-generating trajectories, is modified with the quantification of dynamics difference.
no code implementations • NeurIPS 2021 • Jinxin Liu, Hao Shen, Donglin Wang, Yachen Kang, Qiangxing Tian
Unsupervised reinforcement learning aims to acquire skills without prior goal representations, where an agent automatically explores an open-ended environment to represent goals and learn the goal-conditioned policy.
no code implementations • 12 Nov 2021 • Donglin Wang, Qiuheng Zhou, Sanket Partani, Anjie Qiu, Hans D. Schotten
Nowadays mobile communication is growing fast in the 5G communication industry.
no code implementations • 25 Sep 2019 • Qiangxing Tian, Jinxin Liu, Donglin Wang
By maximizing an information theoretic objective, a few recent methods empower the agent to explore the environment and learn useful skills without supervision.
no code implementations • 2 Mar 2022 • Qingfeng Yao, Jilong Wan, Shuyu Yang, Cong Wang, Linghan Meng, Qifeng Zhang, Donglin Wang
Due to their ability to adapt to different terrains, quadruped robots have drawn much attention in the research field of robot learning.
no code implementations • 2 Mar 2022 • Qingfeng Yao, Jilong Wang, Shuyu Yang, Cong Wang, Hongyin Zhang, Qifeng Zhang, Donglin Wang
The deep learning model extracts key points during animal motion from videos.
no code implementations • ICLR 2022 • Jinxin Liu, Hongyin Zhang, Donglin Wang
Specifically, DARA emphasizes learning from those source transition pairs that are adaptive for the target environment and mitigates the offline dynamics shift by characterizing state-action-next-state pairs instead of the typical state-action distribution sketched by prior offline RL methods.
1 code implementation • 22 Aug 2022 • Siteng Huang, Qiyao Wei, Donglin Wang
Compositional zero-shot learning (CZSL) refers to recognizing unseen compositions of known visual primitives, which is an essential ability for artificial intelligence systems to learn and understand the world.
no code implementations • 13 Sep 2022 • Feng Zhao, Ziqi Zhang, Donglin Wang
This is the first study that we are aware of that looks into dynamic KSG for skill retrieval and learning.
no code implementations • 15 Sep 2022 • Ziqi Zhang, Yile Wang, Yue Zhang, Donglin Wang
Experimental results show that our RL pre-trained models can give close performance compared with the models using the LM training objective, showing that there exist common useful features across these two modalities.
no code implementations • 11 Jan 2023 • Ting Wang, Zongkai Wu, Feiyu Yao, Donglin Wang
First, we propose an Environment Representation Graph (ERG) through object detection to express the environment in semantic level.
no code implementations • 12 Mar 2023 • Min Zhang, Zifeng Zhuang, Zhitao Wang, Donglin Wang, Wenbin Li
OOD exacerbates inconsistencies in magnitudes and directions of task gradients, which brings challenges for GBML to optimize the meta-knowledge by minimizing the sum of task gradients in each minibatch.
no code implementations • 23 May 2023 • Sibo Gai, Donglin Wang, Li He
In this paper, we formulate a new setting, continual offline reinforcement learning (CORL), where an agent learns a sequence of offline reinforcement learning tasks and pursues good performance on all learned tasks with a small replay buffer without exploring any of the environments of all the sequential tasks.
no code implementations • 23 Jun 2023 • Jinxin Liu, Lipeng Zu, Li He, Donglin Wang
As a remedy for the labor-intensive labeling, we propose to endow offline RL tasks with a few expert data and utilize the limited expert data to drive intrinsic rewards, thus eliminating the need for extrinsic rewards.
no code implementations • 26 Jul 2023 • Donglin Wang, Pranav Balasaheb Mohite, Qiuheng Zhou, Anjie Qiu, Hans D. Schotten
To meet the stringent reliability and latency requirements of C-V2X communication, we suggest and assess a retransmission scheme along with a scheme that incorporates varying resource allocations for retransmission in NR V2X communication.
no code implementations • 13 Sep 2023 • Hongyin Zhang, Shuyu Yang, Donglin Wang
To facilitate the development of ORL, we benchmarked 11 ORL algorithms in the realistic quadrupedal locomotion dataset.
no code implementations • 1 Oct 2023 • Teng Xiao, Donglin Wang
This paper studies the problem of learning interactive recommender systems from logged feedbacks without any exploration in online environments.
no code implementations • 7 Oct 2023 • Ziqi Zhang, Xiao Xiong, Zifeng Zhuang, Jinxin Liu, Donglin Wang
Offline-to-online RL can make full use of pre-collected offline datasets to initialize policies, resulting in higher sample efficiency and better performance compared to only using online algorithms alone for policy training.
no code implementations • 10 Nov 2023 • Hongyin Zhang, Diyuan Shi, Zifeng Zhuang, Han Zhao, Zhenyu Wei, Feng Zhao, Sibo Gai, Shangke Lyu, Donglin Wang
Developing robotic intelligent systems that can adapt quickly to unseen wild situations is one of the critical challenges in pursuing autonomous robotics.
no code implementations • 27 Nov 2023 • Siteng Huang, Biao Gong, Yutong Feng, Xi Chen, Yuqian Fu, Yu Liu, Donglin Wang
Experimental results show that existing subject-driven customization methods fail to learn the representative characteristics of actions and struggle in decoupling actions from context features, including appearance.
no code implementations • 28 Nov 2023 • Donglin Wang, Yann Nana Nganso, Hans D. Schotten
We are on the verge of a new age of linked autonomous cars with unheard-of user experiences, dramatically improved air quality and road safety, extremely varied transportation settings, and a plethora of cutting-edge apps.
no code implementations • NeurIPS 2023 • Jinxin Liu, Hongyin Zhang, Zifeng Zhuang, Yachen Kang, Donglin Wang, Bin Wang
Naturally, such a paradigm raises three core questions that are not fully answered by prior non-iterative offline RL counterparts like reward-conditioned policy: (q1) What information should we transfer from the inner-level to the outer-level?
no code implementations • 12 Dec 2023 • Ziqi Zhang, Jingzehua Xu, Zifeng Zhuang, Jinxin Liu, Donglin Wang, Shuai Zhang
Different from previous clipping approaches, we consider increasing the maximum cumulative Return in reinforcement learning (RL) tasks as the preference of the RL task, and propose a bi-level proximal policy optimization paradigm, which involves not only optimizing the policy but also dynamically adjusting the clipping bound to reflect the preference of the RL tasks to further elevate the training outcomes and stability of PPO.
no code implementations • 22 Dec 2023 • Pengxiang Ding, Han Zhao, Zhitao Wang, Zhenyu Wei, Shangke Lyu, Donglin Wang
Within this framework, a notable challenge lies in aligning fine-grained instructions with visual perception information.
no code implementations • 29 Jan 2024 • Ziqi Zhang, Jingzehua Xu, Jinxin Liu, Zifeng Zhuang, Donglin Wang
On the other hand, Decision Transformer (DT) abstracts the decision-making as sequence modeling, showcasing competitive performance on offline RL benchmarks, however, recent studies demonstrate that DT lacks of stitching capability, thus exploit stitching capability for DT is vital to further improve its performance.
no code implementations • 20 Mar 2024 • Wenxuan Song, Han Zhao, Pengxiang Ding, Can Cui, Shangke Lyu, Yaning Fan, Donglin Wang
Multi-task robot learning holds significant importance in tackling diverse and complex scenarios.