Picture for Hao Wen

Hao Wen

BudgetThinker: Empowering Budget-aware LLM Reasoning with Control Tokens

Add code
Aug 24, 2025
Viaarxiv icon

DyCrowd: Towards Dynamic Crowd Reconstruction from a Large-scene Video

Add code
Aug 18, 2025
Viaarxiv icon

LLM-Explorer: Towards Efficient and Affordable LLM-based Exploration for Mobile Apps

Add code
May 15, 2025
Viaarxiv icon

Time's Up! An Empirical Study of LLM Reasoning Ability Under Output Length Constraint

Add code
Apr 22, 2025
Viaarxiv icon

COST: Contrastive One-Stage Transformer for Vision-Language Small Object Tracking

Add code
Apr 02, 2025
Viaarxiv icon

AnyCharV: Bootstrap Controllable Character Video Generation with Fine-to-Coarse Guidance

Add code
Feb 12, 2025
Viaarxiv icon

Improving Skeleton-based Action Recognition with Interactive Object Information

Add code
Jan 09, 2025
Viaarxiv icon

AutoDroid-V2: Boosting SLM-based GUI Agents via Code Generation

Add code
Dec 24, 2024
Viaarxiv icon

IDOL: Instant Photorealistic 3D Human Creation from a Single Image

Add code
Dec 19, 2024
Figure 1 for IDOL: Instant Photorealistic 3D Human Creation from a Single Image
Figure 2 for IDOL: Instant Photorealistic 3D Human Creation from a Single Image
Figure 3 for IDOL: Instant Photorealistic 3D Human Creation from a Single Image
Figure 4 for IDOL: Instant Photorealistic 3D Human Creation from a Single Image
Viaarxiv icon

ChainStream: An LLM-based Framework for Unified Synthetic Sensing

Add code
Dec 13, 2024
Viaarxiv icon