Steerable Visual Representations

• URL: http://arxiv.org/abs/2604.02327v1
• Date: Thu, 02 Apr 2026 17:59:49 GMT
• Title: Steerable Visual Representations
• Abstract summary: We introduce Steerable Visual Representations, whose global and local features can be steered with natural language.<n>Our method also matches or outperforms dedicated approaches on anomaly detection and personalized object discrimination.
• Score: 72.39044430620977
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Pretrained Vision Transformers (ViTs) such as DINOv2 and MAE provide generic image features that can be applied to a variety of downstream tasks such as retrieval, classification, and segmentation. However, such representations tend to focus on the most salient visual cues in the image, with no way to direct them toward less prominent concepts of interest. In contrast, Multimodal LLMs can be guided with textual prompts, but the resulting representations tend to be language-centric and lose their effectiveness for generic visual tasks. To address this, we introduce Steerable Visual Representations, a new class of visual representations, whose global and local features can be steered with natural language. While most vision-language models (e.g., CLIP) fuse text with visual features after encoding (late fusion), we inject text directly into the layers of the visual encoder (early fusion) via lightweight cross-attention. We introduce benchmarks for measuring representational steerability, and demonstrate that our steerable visual features can focus on any desired objects in an image while preserving the underlying representation quality. Our method also matches or outperforms dedicated approaches on anomaly detection and personalized object discrimination, exhibiting zero-shot generalization to out-of-distribution tasks.

Related papers

• ViQ: Text-Aligned Visual Quantized Representations at Any Resolution [91.46185855575789]
  We present ViQ, a Visual Quantized Representations framework, which is designed to balance semantics and details in discrete representations.<n>Our approach structures quantization learning into two stages: text-aligned pre-training and feature discretization.<n>Experiments on multimodal tasks demonstrate that ViQ achieves competitive performance compared to state-of-the-art multimodal vision encoders.
  arXiv  Detail & Related papers  (2026-06-25T17:29:27Z)
• LaCoVL-FER: Landmark-Guided Contrastive Learning Network with Vision-Language Enhancement for Facial Expression Recognition [51.70817823155725]
  We propose a novel landmark-guided contrastive learning network with vision-language enhancement for Facial Expression Recognition (FER)<n>LaCoVL-FER integrates geometric priors from facial landmarks and semantic priors from a vision-language model.<n>Experiments demonstrate that LaCoVL-FER outperforms state-of-the-art methods on three representative real-world FER datasets.
  arXiv  Detail & Related papers  (2026-05-19T13:15:41Z)
• FocalLens: Instruction Tuning Enables Zero-Shot Conditional Image Representations [40.27064688727896]
  We introduce FocalLens, a conditional visual encoding method that produces different representations for the same image based on the context of interest.<n>We show FocalLens leads to performance improvements on a range of downstream tasks including image-image retrieval, image classification, and image-text retrieval.
  arXiv  Detail & Related papers  (2025-04-11T09:07:05Z)
• Beyond Intermediate States: Explaining Visual Redundancy through Language [7.275188652473603]
  Multi-modal Large Langue Models (MLLMs) often process thousands of visual tokens.<n>Visual tokens with low ViT-[cls] association and low text-to-image attention scores can contain recognizable information.<n>We develop a more reliable method for identifying and pruning redundant visual tokens.
  arXiv  Detail & Related papers  (2025-03-26T13:38:10Z)
• Lyrics: Boosting Fine-grained Language-Vision Alignment and Comprehension via Semantic-aware Visual Objects [11.117055725415446]
  Large Vision Language Models (LVLMs) have demonstrated impressive zero-shot capabilities in various vision-language dialogue scenarios. The absence of fine-grained visual object detection hinders the model from understanding the details of images, leading to irreparable visual hallucinations and factual errors. We propose Lyrics, a novel multi-modal pre-training and instruction fine-tuning paradigm that bootstraps vision-language alignment from fine-grained cross-modal collaboration.
  arXiv  Detail & Related papers  (2023-12-08T09:02:45Z)
• Unified Language-Vision Pretraining in LLM with Dynamic Discrete Visual Tokenization [52.935150075484074]
  We introduce a well-designed visual tokenizer to translate the non-linguistic image into a sequence of discrete tokens like a foreign language. The resulting visual tokens encompass high-level semantics worthy of a word and also support dynamic sequence length varying from the image. This unification empowers LaVIT to serve as an impressive generalist interface to understand and generate multi-modal content simultaneously.
  arXiv  Detail & Related papers  (2023-09-09T03:01:38Z)
• Linguistic Query-Guided Mask Generation for Referring Image Segmentation [10.130530501400079]
  Referring image segmentation aims to segment the image region of interest according to the given language expression. We propose an end-to-end framework built on transformer to perform Linguistic query-Guided mask generation.
  arXiv  Detail & Related papers  (2023-01-16T13:38:22Z)
• Visually-augmented pretrained language models for NLP tasks without images [77.74849855049523]
  Existing solutions often rely on explicit images for visual knowledge augmentation. We propose a novel textbfVisually-textbfAugmented fine-tuning approach. Our approach can consistently improve the performance of BERT, RoBERTa, BART, and T5 at different scales.
  arXiv  Detail & Related papers  (2022-12-15T16:13:25Z)
• SgVA-CLIP: Semantic-guided Visual Adapting of Vision-Language Models for Few-shot Image Classification [84.05253637260743]
  We propose a new framework, named Semantic-guided Visual Adapting (SgVA), to extend vision-language pre-trained models. SgVA produces discriminative task-specific visual features by comprehensively using a vision-specific contrastive loss, a cross-modal contrastive loss, and an implicit knowledge distillation. State-of-the-art results on 13 datasets demonstrate that the adapted visual features can well complement the cross-modal features to improve few-shot image classification.
  arXiv  Detail & Related papers  (2022-11-28T14:58:15Z)
• VLMAE: Vision-Language Masked Autoencoder [21.97700040013084]
  We propose a vision-language masked autoencoder framework (VLMAE) for vision-language pre-training. VLMAE employs visual generative learning, facilitating the model to acquire fine-grained and unbiased features.
  arXiv  Detail & Related papers  (2022-08-19T14:39:18Z)
• Two-stage Visual Cues Enhancement Network for Referring Image Segmentation [89.49412325699537]
  Referring Image (RIS) aims at segmenting the target object from an image referred by one given natural language expression. In this paper, we tackle this problem by devising a Two-stage Visual cues enhancement Network (TV-Net) Through the two-stage enhancement, our proposed TV-Net enjoys better performances in learning fine-grained matching behaviors between the natural language expression and image.
  arXiv  Detail & Related papers  (2021-10-09T02:53:39Z)
• Self-Supervised Representation Learning from Flow Equivariance [97.13056332559526]
  We present a new self-supervised learning representation framework that can be directly deployed on a video stream of complex scenes. Our representations, learned from high-resolution raw video, can be readily used for downstream tasks on static images.
  arXiv  Detail & Related papers  (2021-01-16T23:44:09Z)
• Learning Visual Representations with Caption Annotations [19.24013129952071]
  We propose a proxy task to learn visual representations over image-caption pairs. ICMLM consists in predicting masked words in captions by relying on visual cues. Our experiments confirm that image captions can be leveraged to inject global and localized semantic information into visual representations.
  arXiv  Detail & Related papers  (2020-08-04T08:04:16Z)

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.