9 papers
When Policy Entropy Constraint Fails: Preserving Diversity in Flow-based RLHF via Perceptual Entropy
Xiaofeng Tan, Jun Liu, Bin-Bin Gao +5
RLHF is widely used to align flow-matching text-to-image models with human preferences, but often leads to severe diversity collapse after fine-tuning. In RL, diversity is often as…
MotionRFT: Unified Reinforcement Fine-Tuning for Text-to-Motion Generation
Xiaofeng Tan, Wanjiang Weng, Hongsong Wang +3
Text-to-motion generation has advanced with diffusion- and flow-based generative models, yet supervised pretraining remains insufficient to align models with high-level objectives…
Bilingual Text-to-Motion Generation: A New Benchmark and Baselines
Wanjiang Weng, Xiaofeng Tan, Xiangbo Shu +3
Text-to-motion generation holds significant potential for cross-linguistic applications, yet it is hindered by the lack of bilingual datasets and the poor cross-lingual semantic un…
EasyTune: Efficient Step-Aware Fine-Tuning for Diffusion-Based Motion Generation
Xiaofeng Tan, Wanjiang Weng, Haodong Lei +1
In recent years, motion generative models have undergone significant advancement, yet pose challenges in aligning with downstream objectives. Recent studies have shown that using d…
ConsistentRFT: Reducing Visual Hallucinations in Flow-based Reinforcement Fine-Tuning
Xiaofeng Tan, Jun Liu, Yuanting Fan +7
Reinforcement Fine-Tuning (RFT) on flow-based models is crucial for preference alignment. However, they often introduce visual hallucinations like over-optimized details and semant…
ReAlign: Text-to-Motion Generation via Step-Aware Reward-Guided Alignment
Wanjiang Weng, Xiaofeng Tan, Junbo Wang +3
Text-to-motion generation, which synthesizes 3D human motions from text inputs, holds immense potential for applications in gaming, film, and robotics. Recently, diffusion-based me…