AI研究論文デイリー

Building and better understanding vision-language models: insights and future directions

CustomCrafter: Customized Video Generation with Preserving Motion and Concept Composition Abilities

CODE: Confident Ordinary Differential Editing

T3M: Text Guided 3D Human Motion Synthesis from Speech

HiRED: Attention-Guided Token Dropping for Efficient Inference of High-Resolution Vision-Language Models in Resource-Constrained Environments

A Web-Based Solution for Federated Learning with LLM-Based Automation

MME-RealWorld: Could Your Multimodal LLM Challenge High-Resolution Real-World Scenarios that are Difficult for Humans?

Multi-Layer Transformers Gradient Can be Approximated in Almost Linear Time

LayerPano3D: Layered 3D Panorama for Hyper-Immersive Scene Generation

FLoD: Integrating Flexible Level of Detail into 3D Gaussian Splatting for Customizable Rendering

Memory-Efficient LLM Training with Online Subspace Descent

RoundTable: Leveraging Dynamic Schema and Contextual Autocomplete for Enhanced Query Precision in Tabular Question Answering