Locality Alignment Improves Vision-Language Models

• URL: http://arxiv.org/abs/2410.11087v1
• Date: Mon, 14 Oct 2024 21:01:01 GMT
• Title: Locality Alignment Improves Vision-Language Models
• Authors: Ian Covert, Tony Sun, James Zou, Tatsunori Hashimoto,
• Abstract summary: Vision language models (VLMs) have seen growing adoption in recent years, but many still struggle with basic spatial reasoning errors. We propose a new efficient post-training stage for ViTs called locality alignment. We show that locality-aligned backbones improve performance across a range of benchmarks.
• Score: 55.275235524659905
• License:
• Abstract: Vision language models (VLMs) have seen growing adoption in recent years, but many still struggle with basic spatial reasoning errors. We hypothesize that this is due to VLMs adopting pre-trained vision backbones, specifically vision transformers (ViTs) trained with image-level supervision and minimal inductive biases. Such models may fail to encode the class contents at each position in the image, and our goal is to resolve this by ensuring that the vision backbone effectively captures both local and global image semantics. Our main insight is that we do not require new supervision to learn this capability -- pre-trained models contain significant knowledge of local semantics that we can extract and use for scalable self-supervision. We propose a new efficient post-training stage for ViTs called locality alignment and a novel fine-tuning procedure called MaskEmbed that uses a masked reconstruction loss to learn semantic contributions for each image patch. We first evaluate locality alignment with a vision-only benchmark, finding that it improves a model's performance at a patch-level semantic segmentation task, especially for strong backbones trained with image-caption pairs (e.g., CLIP and SigLIP). We then train a series of VLMs with and without locality alignment, and show that locality-aligned backbones improve performance across a range of benchmarks, particularly ones that involve spatial understanding (e.g., RefCOCO, OCID-Ref, TallyQA, VSR, AI2D). Overall, we demonstrate that we can efficiently learn local semantic extraction via a locality alignment stage, and that this procedure complements existing VLM training recipes that use off-the-shelf vision backbones.

Related papers

• Context-Based Visual-Language Place Recognition [4.737519767218666]
  A popular approach to vision-based place recognition relies on low-level visual features. We introduce a novel VPR approach that remains robust to scene changes and does not require additional training. Our method constructs semantic image descriptors by extracting pixel-level embeddings using a zero-shot, language-driven semantic segmentation model.
  arXiv  Detail & Related papers  (2024-10-25T06:59:11Z)
• Contrastive Localized Language-Image Pre-Training [60.4967533101887]
  Contrastive Language-Image Pre-training (CLIP) has been a celebrated method for training vision encoders to generate image/text representations. We propose Contrastive Localized Language-Image Pre-training (CLOC) by complementing CLIP with region-text contrastive loss and modules. CLOC enables high-quality regional embeddings for image region recognition and retrieval tasks.
  arXiv  Detail & Related papers  (2024-10-03T17:56:09Z)
• Harnessing Vision-Language Pretrained Models with Temporal-Aware Adaptation for Referring Video Object Segmentation [34.37450315995176]
  Current Referring Video Object (RVOS) methods typically use vision and language models pretrained independently as backbones. We propose a temporal-aware prompt-tuning method, which adapts pretrained representations for pixel-level prediction. Our method performs favorably against state-of-the-art algorithms and exhibits strong generalization abilities.
  arXiv  Detail & Related papers  (2024-05-17T08:14:22Z)
• Grounding Everything: Emerging Localization Properties in Vision-Language Transformers [51.260510447308306]
  We show that pretrained vision-language (VL) models allow for zero-shot open-vocabulary object localization without any fine-tuning. We propose a Grounding Everything Module (GEM) that generalizes the idea of value-value attention introduced by CLIPSurgery to a self-self attention path. We evaluate the proposed GEM framework on various benchmark tasks and datasets for semantic segmentation.
  arXiv  Detail & Related papers  (2023-12-01T19:06:12Z)
• Contrastive Vision-Language Alignment Makes Efficient Instruction Learner [31.281236193979165]
  We study the task of extending the large language model (LLM) into a vision-language instruction-following model. Existing methods typically train a visual adapter to align the representation between a pre-trained vision transformer (ViT) and the LLM by a generative image captioning loss. We propose CG-VLM that applies Contrastive and Generative alignment objectives to effectively align the representation of ViT and LLM.
  arXiv  Detail & Related papers  (2023-11-29T03:29:46Z)
• IFSeg: Image-free Semantic Segmentation via Vision-Language Model [67.62922228676273]
  We introduce a novel image-free segmentation task where the goal is to perform semantic segmentation given only a set of the target semantic categories. We construct this artificial training data by creating a 2D map of random semantic categories and another map of their corresponding word tokens. Our model not only establishes an effective baseline for this novel task but also demonstrates strong performances compared to existing methods.
  arXiv  Detail & Related papers  (2023-03-25T08:19:31Z)
• Location-Aware Self-Supervised Transformers [74.76585889813207]
  We propose to pretrain networks for semantic segmentation by predicting the relative location of image parts. We control the difficulty of the task by masking a subset of the reference patch features visible to those of the query. Our experiments show that this location-aware pretraining leads to representations that transfer competitively to several challenging semantic segmentation benchmarks.
  arXiv  Detail & Related papers  (2022-12-05T16:24:29Z)
• Pretraining the Vision Transformer using self-supervised methods for vision based Deep Reinforcement Learning [0.0]
  We study pretraining a Vision Transformer using several state-of-the-art self-supervised methods and assess the quality of the learned representations. Our results show that all methods are effective in learning useful representations and avoiding representational collapse. The encoder pretrained with the temporal order verification task shows the best results across all experiments.
  arXiv  Detail & Related papers  (2022-09-22T10:18:59Z)
• PGL: Prior-Guided Local Self-supervised Learning for 3D Medical Image Segmentation [87.50205728818601]
  We propose a PriorGuided Local (PGL) self-supervised model that learns the region-wise local consistency in the latent feature space. Our PGL model learns the distinctive representations of local regions, and hence is able to retain structural information.
  arXiv  Detail & Related papers  (2020-11-25T11:03:11Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.