arXiv AI

Aligning Forest and Trees in Images & Long Captions for Visually Grounded Understanding

The paper introduces CAFT, a vision‑language model that improves understanding of long, detail‑rich captions by aligning local text regions with image parts before forming a global representation. It uses a fine‑to‑coarse image encoder and a part‑to‑whole text encoder to discover localized semantics and compose them into a holistic image‑text embedding. Trained on 30 M image‑text pairs, CAFT outperforms existing methods on six long‑text retrieval benchmarks and scales well, learning fine‑grained representations without explicit region supervision.

arXiv AI
Aug 24

CFM: Language-aligned Concept Foundation Model for Vision

The paper introduces CFM, a language‑aligned concept foundation model for vision that generates fine‑grained, human‑interpretable concepts with spatial grounding. By pairing CFM with a strong semantic foundation model, it provides explanations for downstream tasks such as classification, segmentation, and captioning. The authors also analyze local co‑occurrence of concepts to define relationships, improving concept naming and yielding richer explanations while maintaining competitive performance.

By Kai Wittenmayer, Sukrut Rao, Amin Parchami-Araghi, Bernt Schiele, Jonas Fischer
arXiv Computer Vision
Oct 16

Flow-CLIP: Image-Conditioned Residual Scoring for CompositionalImage-Text Matching

Flow-CLIP introduces a method to improve compositional image-text matching by refining global image-text alignment with image-conditioned changes in token–patch evidence. It combines Global Compatibility Calibration (GCC) for sentence-level similarity, Image-Conditioned Evidence Refinement (ICER) to incorporate visual patch context, and Cross-Image Margin Ranking (CIMR) to enhance ranking margins when images support correct descriptions. Experiments demonstrate that Flow-CLIP outperforms pretrained CLIP and other global alignment approaches on binding accuracy and reduces reliance on language priors.

By Minghong Wei, Enguang Zhu, Pu Cao, Lu Yang, Qing Song
Hugging Face Trending Papers
Jul 15

Fine-grained CLIP fine-tuning with self-annotated region alignment

Contrastive Language-Image Pre-training (CLIP) has been shown to have limitations in its fine-grained dense feature representation, due to its pre-training focusing on matching the whole image to a text description. Considering the large data and computational burden in pre-training a vision-language model from scratch, a series of works aim to enhance the fine-grained ability of CLIP through a fine-tuning scheme.

arXiv Computation and Language
Sep 17

PANORAMA: Panoptic Grounded Captioning via Mask Proposal Selection

PANORAMA introduces a new panoptic grounded captioning framework that jointly generates detailed image captions and associates each phrase with precise pixel-level masks. The authors create PanoCaps, a human‑annotated benchmark with dense captions and near‑complete pixel coverage, and propose a phrase‑mask matching protocol with a generalized Panoptic Quality metric. PANORAMA conditions a pretrained segmenter on contextualized phrase representations, learns to select appropriate masks, and achieves state‑of‑the‑art grounding performance on PanoCaps and other pixel‑level tasks.

By Sara Pieri, Evangelos Kazakos, Shizhe Chen, Josef Sivic, Cordelia Schmid
arXiv Computer Vision
Aug 28

SOCO: Benchmarking Semantic Object Correspondence in Vision Foundation Models

SOCO is a new benchmark for Semantic Object Correspondence that introduces a taxonomy of correspondence types and provides consistent, functionally meaningful keypoint annotations across 100 categories and over 1M correspondence pairs. It also includes keypoint language descriptions, enabling evaluation of large vision‑language models and their fine‑grained part‑level understanding. Experiments show that vision foundation backbones encode strong semantic structure but transfer correspondences poorly across related categories, LVLMs excel at text‑prompted part localization but lag in visual‑reference matching, and correspondence performance predicts dense downstream tasks more strongly than ImageNet classification.

By Olaf D\"unkel, Basavaraj Sunagad, Haoran Wang, David T. Hoffmann, Christian Theobalt, Adam Kortylewski
arXiv Computer Vision
Nov 9

UniMore: Fine-Grained Supervision for Unified Understanding and Generation

UniMore is a post‑training technique that enhances unified multimodal models by adding fine‑grained supervision derived from image–text pairs. It introduces two complementary learning tasks: a generation route that predicts grounded image crops from phrases, and an understanding route that ranks captions based on detail fidelity. These tasks are optimized separately and then merged into the original model, improving performance on both image generation and captioning benchmarks without altering inference.

By Haoyue Bai, Yinyi Luo, Yicheng He, Qingsong Wen, Sharon Li, Jindong Wang