task-specific fine-tuning
- ESCA: Contextualizing Embodied Agents via Scene-Graph Generation
- FOCUS: Internal MLLM Representations for Efficient Fine-Grained Visual Question Answering
- Geometry-Aware Collaborative Multi-Solutions Optimizer for Model Fine-Tuning with Parameter Efficiency
- MomentSeeker: A Task-Oriented Benchmark For Long-Video Moment Retrieval
- RoomEditor: High-Fidelity Furniture Synthesis with Parameter-Sharing U-Net
- Zero-Shot Detection of LLM-Generated Text via Implicit Reward Model