ZB
Zechen Bai
cs.CVcs.AIcs.ROcs.CLcs.LGcs.HCcs.MMcs.IRcs.MA
On Valency
published · living versionsW_jetaqt9m·v1 · currentpublished
Multimodal Embodied Agents: A Unified Capability-centric Perspective from Computer-Use to Robot-Use
with Yanzhe Chen, Ziyi Yang, Jifeng Zhu, Qiming Huang +12
1 version
Preprints & journals
24 papers in the corpus · 2020–2026Toward Recursive Self-Improvement in Physical AI2610.29677v1 · Meng Luo, Zechen Bai, Shanqing Xu et al.2026 · 0 citationsarXivon Valency
Multimodal Embodied Agents: A Unified Capability-centric Perspective from Computer-Use to Robot-Use2610.30108v1 · Yanzhe Chen, Ziyi Yang, Jifeng Zhu et al.2026 · 0 citationsarXivon Valency
Show-Harness: Just a VLM Agent Can Play Robots2609.10522v1 · Yanzhe Chen, Zechen Bai, Zhijun Cao et al.2026 · 0 citationsarXiv
Demo2Tutorial: From Human Experience to Multimodal Software Tutorials2606.03951v1 · Zechen Bai, Zhiheng Chen, Yiqi Lin et al.2026 · 0 citationsarXiv
Escaping the Diversity Trap in Robotic Manipulation via Anchor-Centric Adaptation2605.07381v1 · Yanzhe Chen, Kevin Yuchen Ma, Qi Lv et al.2026 · 0 citationsarXiv
Kiwi-Edit: Versatile Video Editing via Instruction and Reference Guidance2603.02175v4 · Yiqi Lin, Guoqiang Liang, Ziyun Zeng et al.2026 · 0 citationsarXiv
SIGMA: Selective-Interleaved Generation with Multi-Attribute Tokens2602.07564v1 · Xiaoyan Zhang, Zechen Bai, Haofan Wang et al.2026 · 0 citationsarXiv
World-VLA-Loop: Closed-Loop Learning of Video World Model and VLA Policy2602.06508v2 · Xiaokang Liu, Zechen Bai, Hai Ci et al.2026 · 0 citationsarXiv
EVOLVE-VLA: Test-Time Training from Environment Feedback for Vision-Language-Action Models2512.14666v1 · Zechen Bai, Chen Gao, Mike Zheng Shou2025 · 0 citationsarXiv
Impossible Videos2503.14378v1 · Zechen Bai, Hai Ci, Mike Zheng Shou2025 · 0 citationsarXiv
ShowUI: One Vision-Language-Action Model for GUI Visual Agent2411.17465v1 · Kevin Qinghong Lin, Linjie Li, Difei Gao et al.2024 · 32 citationsarXiv
Factorized Visual Tokenization and Generation2411.16681v2 · Zechen Bai, Jianxiong Gao, Ziteng Gao et al.2024 · 0 citationsarXiv
One Token to Seg Them All: Language Instructed Reasoning Segmentation in Videos2409.19603v1 · Zechen Bai, Tong He, Haiyang Mei et al.2024 · 37 citationsarXiv
Bridging Information Asymmetry in Text-video Retrieval: A Data-centric Approach2408.07249v2 · Zechen Bai, Tianjun Xiao, Tong He et al.2024 · 0 citationsarXiv
Adaptive Slot Attention: Object Discovery with Dynamic Slot Number2406.09196v1 · Ke Fan, Zechen Bai, Tianjun Xiao et al.2024 · 10 citationsarXiv
LOVA3: Learning to Visual Question Answering, Asking and Assessment2405.14974v3 · Henry Hengyuan Zhao, Pan Zhou, Difei Gao et al.2024 · 1 citationarXiv
Bring Your Own Character: A Holistic Solution for Automatic Facial Animation Generation of Customized Characters2402.13724v1 · Zechen Bai, Peng Chen, Xiaolan Peng et al.2024 · 8 citationsarXiv
Skip \n: A Simple Method to Reduce Hallucination in Large Vision-Language Models2402.01345v6 · Zongbo Han, Zechen Bai, Haiyang Mei et al.2024 · 0 citationsarXiv
ASSISTGUI: Task-Oriented Desktop Graphical User Interface Automation2312.13108v2 · Difei Gao, Lei Ji, Zechen Bai et al.2023 · 0 citationsarXiv
Unsupervised Open-Vocabulary Object Localization in Videos2309.09858v2 · Ke Fan, Zechen Bai, Tianjun Xiao et al.2023 · 6 citationsarXiv
Object-Centric Multiple Object Tracking2309.00233v2 · Zixu Zhao, Jiaze Wang, Max Horn et al.2023 · 13 citationsarXiv
Explain Me the Painting: Multi-Topic Knowledgeable Art Description Generation2109.05743v1 · Zechen Bai, Yuta Nakashima, Noa Garcia2021 · 46 citationsarXiv
Unsupervised Multi-Source Domain Adaptation for Person Re-Identification2104.12961v1 · Zechen Bai, Zhigang Wang, Jian Wang et al.2021 · 106 citationsarXiv
Show, Recall, and Tell: Image Captioning with Recall Mechanism2001.05876v3 · Li Wang, Zechen Bai, Yonghua Zhang et al.2020 · 62 citationsarXiv
Career total: 36 works. 24 are in this corpus.
Profile built from the corpus for this byline.
Author records are still filling in while the Hub is in alpha. If this is your page, you'll be able to claim it soon. Spot a mistake? Tell us.