CAPability: A Comprehensive Visual Caption Benchmark for Evaluating Both Correctness and Thoroughness

ReMatch has been introduced as a new framework that utilizes the generative capabilities of Multimodal Large Language Models (MLLMs) for enhanced multimodal retrieval. This approach trains the MLLM end-to-end, employing a chat-style generative matching stage that assesses relevance from various inputs, including raw data and projected embeddings.

Restora-Flow has been introduced as a training-free method for image restoration that utilizes flow matching sampling guided by a degradation mask. This innovative approach aims to enhance the quality of image restoration tasks such as inpainting, super-resolution, and denoising while addressing the long processing times and over-smoothing issues faced by existing methods.

RobustMerge has been introduced as a parameter-efficient model merging method designed for multi-task learning in machine learning language models (MLLMs), emphasizing direction robustness during the merging process. This approach addresses the challenges of merging expert models without data leakage, which has become increasingly important as model sizes and data complexity grow.

The recent introduction of EmoFeedback$^2$ aims to enhance continuous emotional image generation (C-EICG) by utilizing a large vision-language model (LVLM) to provide reward and textual feedback, addressing the limitations of existing methods that struggle with emotional continuity and fidelity. This paradigm allows for better alignment of generated images with user emotional descriptions.

A new study has introduced a diffusion-based inpainting model adapted for image layer decomposition, addressing the challenges of separating images into distinct layers for independent editing. This model employs lightweight finetuning and a multi-modal context fusion module to enhance detail preservation in the latent space, achieving superior results in object removal and occlusion recovery using a synthetic dataset.

A new benchmark called CaptionQA has been introduced to evaluate the utility of model-generated captions in supporting downstream tasks across various domains, including Natural, Document, E-commerce, and Embodied AI. This benchmark consists of 33,027 annotated multiple-choice questions that require visual information to answer, aiming to assess whether captions can effectively replace images in multimodal systems.

A novel framework for Trustworthy Multi-View Classification (TMVC) has been proposed, addressing the challenges of reliable decision-making in scenarios with heterogeneous and conflicting multi-source information. This framework introduces prototypes to represent neighbor structures of each view, simplifying the learning of intra-view relations and enhancing consistency across inter-view relationships.

LLaVA-UHD v3 has been introduced as a new multi-modal large language model (MLLM) that utilizes Progressive Visual Compression (PVC) for efficient native-resolution encoding, enhancing visual understanding capabilities while addressing computational overhead. This model integrates refined patch embedding and windowed token compression to optimize performance in vision-language tasks.