DirectSpeech2LLM: A Simple End-to-End Framework to Mitigate Prompt Overfitting in Speech-LLMs
Research
Research News
New papers, breakthroughs and academic AI research
Later Is Better: Token Reduction for ViTs Under Distribution Shift
Research
Foresight-over-Graph: Reasoning Beyond Local Horizons for Knowledge Base Question Answering
Research
Structure-aware Keypoint Localization for Videofluoroscopic Swallowing Study
Research
From Laboratory to Road: Evaluating Wearable Gaze Accuracy for Driving
Research
Geometry-Constrained Bidirectional Point Cloud Registration for Thin, Sheet-Like Heritage Artifacts
Research
Rethinking Cross-Tokenizer On-Policy Distillation: From Alignment Coverage to Supervision Reliability
Research
SAGE: Semantic Anchor-Guided Evolution for Grounded Medical QA Data Synthesis
Research
Penalty-Framed No-Valid-Option MCQA: Analyzing LLM Abstention under Invalid Choices
Research
STRUCTURALCOST: A controlled reading time dataset for modeling human sentence processing difficulty
Research
Foveated Compression: Selective High-Resolution Preservation for Token-Efficient VLMs
Research
Image-Space Refraction Correction for Underwater 3D Reconstruction: Warping Flat-Port Views into Pinhole Perspective
Research