AI Onderzoekspapers Dagelijks

VisionZip: Longer is Better but Not Necessary in Vision Language Models

Structured 3D Latents for Scalable and Versatile 3D Generation

Aguvis: Unified Pure Vision Agents for Autonomous GUI Interaction

Florence-VL: Enhancing Vision-Language Models with Generative Vision Encoder and Depth-Breadth Fusion

NVILA: Efficient Frontier Visual Language Models

Evaluating Language Models as Synthetic Data Generators

Code-as-Monitor: Constraint-aware Visual Programming for Reactive and Proactive Robotic Failure Detection

A Noise is Worth Diffusion Guidance

MV-Adapter: Multi-view Consistent Image Generation Made Easy

AnyDressing: Customizable Multi-Garment Virtual Dressing via Latent Diffusion Models

Negative Token Merging: Image-based Adversarial Feature Guidance

Densing Law of LLMs

Global MMLU: Understanding and Addressing Cultural and Linguistic Biases in Multilingual Evaluation

Infinity: Scaling Bitwise AutoRegressive Modeling for High-Resolution Image Synthesis

HumanEdit: A High-Quality Human-Rewarded Dataset for Instruction-based Image Editing

Personalized Multimodal Large Language Models: A Survey

Monet: Mixture of Monosemantic Experts for Transformers

OmniFlow: Any-to-Any Generation with Multi-Modal Rectified Flows

Towards Universal Soccer Video Understanding

Discriminative Fine-tuning of LVLMs

Marco-LLM: Bridging Languages via Massive Multilingual Training for Cross-Lingual Enhancement

MEMO: Memory-Guided Diffusion for Expressive Talking Video Generation

ZipAR: Accelerating Autoregressive Image Generation through Spatial Locality

KV Shifting Attention Enhances Language Modeling

4Real-Video: Learning Generalizable Photo-Realistic 4D Video Diffusion

p-MoD: Building Mixture-of-Depths MLLMs via Progressive Ratio Decay

Scaling Inference-Time Search with Vision Value Model for Improved Visual Comprehension

MRGen: Diffusion-based Controllable Data Engine for MRI Segmentation towards Unannotated Modalities

SynFinTabs: A Dataset of Synthetic Financial Tables for Information and Table Extraction

Challenges in Trustworthy Human Evaluation of Chatbots

Establishing Task Scaling Laws via Compute-Efficient Model Ladders