Bihan Wen lists you as their PhD student. Claim this profile to confirm it and keep the rest of your record right.
Claim this profileAcademic lineage
View as a treeAdvisors
Works43 from public data
- RL-VLM-F: Reinforcement Learning from Vision Language Foundation Model Feedback200
RL-VLM-F is proposed, a method that automatically generates reward functions for agents to learn new tasks, using only a text description of the task goal and the agent's visual observations, by leveraging feedbacks from vision language foundation models (VLMs).
- ToolFlowNet: Robotic Manipulation with Tools via Predicting Tool Flow from Point Clouds90
This paper proposes a novel framework for learning policies from point clouds for robotic manipulation with tools, using a novel neural network, ToolFlowNet, which predicts dense per-point flow on the tool that the robot controls, and then uses the flow to derive the transformation that the machine should execute.
- One Policy to Dress Them All: Learning to Dress People with Diverse Poses and Garments43
This work develops a robot-assisted dressing system that is able to dress different garments on people with diverse poses from partial point cloud observations, based on a learned policy and uses reinforcement learning to combine multiple policies trained on different ranges of human arm poses into a single policy.
- 39
- Decomposed Guided Dynamic Filters for Efficient RGB-Guided Depth Completion31
This paper proposes to decompose the guided dynamic filters into a spatially-shared component multiplied by content-adaptive adaptors at each spatial location and introduces two decomposition schemes, which decompose the filters by splitting the filter structure and using spatial-wise attention, respectively.
- Real-World Offline Reinforcement Learning from Vision Language Model Feedback30
A novel system that automatically generates reward labels for offline datasets using preference feedback from a vision-language model and a text description of the task is proposed, which demonstrates the system’s applicability to a complex real-world robot-assisted dressing task.
- 23
- 16
- 15
- 15
- 11
- 9
- 8
- Neural Dynamics for Computing and Automation: A Survey5
For each application area, the detailed problems are classified and an analysis of the current neural dynamic methods or models are provided and some interesting research topics and directions for further investigation are identified.
- 5
- 5
- 5
- 4
- 4
- Force-Modulated Visual Policy for Robot-Assisted Dressing with Arm Motions3
This work develops a robot-assisted dressing system capable of handling partial observations with visual occlusions, as well as robustly adapting to arm motions during the dressing process, and greatly outperforms prior baselines in terms of task completion and user feedback.
- 3
- Transcriptome Profiling of a Common Mistletoe Species Parasitizing Four Typical Host Species in Urban Southwest China3
The transcriptomes of T. nigrans individuals parasitizing four typical host species were sequenced, and the expression profiles and metabolic pathways were compared among hosts to provide a foundation for further exploration of the detailed molecular mechanisms involved in plant parasitism.
- An Image Fusion Method Based on Color Space for Multimedia Sensor Networks3
The monitor task of the same scene was assigned to three highly-correlated video sensor nodes based on YUV color space and the view correlation of adjacent nodes and the proposed depth information model was used.
- 2
- Coordinated cyberattacks of microgeneration systems2
A cross-space cascading failure evolution model based on the state-dependent evolution of the cyber events chain and the power events chain is proposed, which aims to prevent computers from providing services to users and to control the cyber system directly.
- Experimental Study of Intelligent Autopilot for Surface Vessels Based on Neural Network Optimised PID Controller2
The experimental results indicate that the proposed RBFNN PID based autopilot system for a surface vessel can decrease the course keeping deviations with reasonable rudder actions.
- Deep Learning for Image Understanding2
A coarse-to-fine LSTM based method is proposed that decomposes the original image description into a skeleton sentence and its notable attributes, and demonstrates that in this way the language model can generate better descriptions, with the capability to generate image descriptions that better accommodates user preference.
- 2
- 1
- 1
- –
- –
- RealisticDreamer: Guidance Score Distillation for Few-shot Gaussian Splatting–
This work introduces an unified guidance form to correct the noise prediction result of VDM, incorporating both a depth warp guidance based on real depth maps and a guidance based on semantic image features, ensuring that the score update direction from VDM aligns with the correct camera pose and accurate geometry.
- Research on Hidden Backdoor Prompt Attack Method–
Findings set a new benchmark for undetectable backdoor attacks and underscore the urgent need for advancements in prompt-based defense strategies.
- –
- –
- –
- –
- –
- –
- –
- Function and Application on The Virtual Studio–
The paper discussed the work principle and the function and the application of the virtual studio, the combination of traditional video technique and virtual reality technique, which is important replace for traditional TV studio.
- Characteristics and Maintenance of Digital Video–
The characteristics and the maintenance of digital video in actual operation are explained.
Publication data from OpenAlex, with missing venues and authors filled in from Crossref; citation counts are the higher of OpenAlex and Semantic Scholar, last synced 2026-10-12. One-sentence summaries under some papers are written by Semantic Scholar’s model. Citation counts may be lower than on Google Scholar, which indexes more sources.
Report an error
Wrong papers, two people merged into one, or a profile that should not be here? Tell us and we will fix or hide it.