Dynamic Prompting of Frozen Text-to-Image Diffusion Models for Panoptic Narrative Grounding

• URL: http://arxiv.org/abs/2409.08251v1
• Date: Thu, 12 Sep 2024 17:48:22 GMT
• Title: Dynamic Prompting of Frozen Text-to-Image Diffusion Models for Panoptic Narrative Grounding
• Authors: Hongyu Li, Tianrui Hui, Zihan Ding, Jing Zhang, Bin Ma, Xiaoming Wei, Jizhong Han, Si Liu,
• Abstract summary: We propose an Extractive-Injective Phrase Adapter (EIPA) bypass within the Diffusion UNet to dynamically update phrase prompts with image features. We also design a Multi-Level Mutual Aggregation (MLMA) module to reciprocally fuse multi-level image and phrase features for segmentation refinement.
• Score: 39.73180294057053
• License: http://creativecommons.org/licenses/by-nc-nd/4.0/
• Abstract: Panoptic narrative grounding (PNG), whose core target is fine-grained image-text alignment, requires a panoptic segmentation of referred objects given a narrative caption. Previous discriminative methods achieve only weak or coarse-grained alignment by panoptic segmentation pretraining or CLIP model adaptation. Given the recent progress of text-to-image Diffusion models, several works have shown their capability to achieve fine-grained image-text alignment through cross-attention maps and improved general segmentation performance. However, the direct use of phrase features as static prompts to apply frozen Diffusion models to the PNG task still suffers from a large task gap and insufficient vision-language interaction, yielding inferior performance. Therefore, we propose an Extractive-Injective Phrase Adapter (EIPA) bypass within the Diffusion UNet to dynamically update phrase prompts with image features and inject the multimodal cues back, which leverages the fine-grained image-text alignment capability of Diffusion models more sufficiently. In addition, we also design a Multi-Level Mutual Aggregation (MLMA) module to reciprocally fuse multi-level image and phrase features for segmentation refinement. Extensive experiments on the PNG benchmark show that our method achieves new state-of-the-art performance.

Related papers

• Retrieve and Segment: Are a Few Examples Enough to Bridge the Supervision Gap in Open-Vocabulary Segmentation? [13.02242959998332]
  Open-vocabulary segmentation (OVS) extends the zero-shot recognition capabilities of vision-language models (VLMs) to pixel-level prediction.<n>We introduce a few-shot setting that augments textual prompts with a support set of pixel-annotated images.<n>We propose a retrieval-augmented test-time adapter that learns a lightweight, per-image classifier by fusing textual and visual support features.
  arXiv  Detail & Related papers  (2026-02-26T18:45:33Z)
• Seg4Diff: Unveiling Open-Vocabulary Segmentation in Text-to-Image Diffusion Transformers [56.76198904599581]
  Text-to-image diffusion models excel at translating language prompts into implicitly grounding concepts through their cross-modal attention mechanisms.<n>Recent multi-modal diffusion transformers extend this by introducing joint self-attentiond image and text tokens, enabling richer and more scalable cross-modal alignment.<n>We introduce Seg4Diff, a systematic framework for analyzing the attention structures of MM-DiT, with a focus on how specific layers propagate semantic information from text to image.
  arXiv  Detail & Related papers  (2025-09-22T17:59:54Z)
• Text4Seg++: Advancing Image Segmentation via Generative Language Modeling [52.07442359419673]
  We propose a novel text-as-mask paradigm that casts image segmentation as a text generation problem.<n>Key innovation is semantic descriptors, a new textual representation of segmentation masks.<n>Experiments on natural and remote sensing datasets show that Text4Seg++ consistently outperforms state-of-the-art models.
  arXiv  Detail & Related papers  (2025-09-08T04:07:14Z)
• GS: Generative Segmentation via Label Diffusion [59.380173266566715]
  Language-driven image segmentation is a fundamental task in vision-language understanding, requiring models to segment regions of an image corresponding to natural language expressions.<n>Recent diffusion models have been introduced to this domain, but existing approaches remain image-centric.<n>We propose GS (Generative Label), a novel framework that formulates segmentation itself as a generative task.<n> Experimental results show that GS significantly outperforms existing discriminative and diffusion-based methods, setting a new state-of-the-art for language-driven segmentation.
  arXiv  Detail & Related papers  (2025-08-27T16:28:15Z)
• InvSeg: Test-Time Prompt Inversion for Semantic Segmentation [33.60580908728705]
  InvSeg is a test-time prompt inversion method for semantic segmentation. We introduce Contrastive Soft Clustering to align masks with the image's structure information. InvSeg learns context-rich text prompts in embedding space and achieves accurate semantic alignment across modalities.
  arXiv  Detail & Related papers  (2024-10-15T10:20:31Z)
• Exploring Phrase-Level Grounding with Text-to-Image Diffusion Model [61.389233691596004]
  We introduce the DiffPNG framework, which capitalizes on the diffusion's architecture for segmentation by decomposing the process into a sequence of localization, segmentation, and refinement steps. Our experiments on the PNG dataset demonstrate that DiffPNG achieves strong performance in the zero-shot PNG task setting.
  arXiv  Detail & Related papers  (2024-07-07T13:06:34Z)
• Open-Vocabulary Attention Maps with Token Optimization for Semantic Segmentation in Diffusion Models [1.6450779686641077]
  We introduce Open-Vocabulary Attention Maps (OVAM)-a training-free method for text-to-image diffusion models. We evaluate these tokens within existing state-of-the-art Stable Diffusion extensions.
  arXiv  Detail & Related papers  (2024-03-21T10:56:12Z)
• Prefix-diffusion: A Lightweight Diffusion Model for Diverse Image Captioning [36.4086473737433]
  We propose a lightweight image captioning network in combination with continuous diffusion, called Prefix-diffusion. To achieve diversity, we design an efficient method that injects prefix image embeddings into the denoising process of the diffusion model. In order to reduce trainable parameters, we employ a pre-trained model to extract image features and further design an extra mapping network.
  arXiv  Detail & Related papers  (2023-09-10T08:55:24Z)
• Energy-Based Cross Attention for Bayesian Context Update in Text-to-Image Diffusion Models [62.603753097900466]
  We present a novel energy-based model (EBM) framework for adaptive context control by modeling the posterior of context vectors. Specifically, we first formulate EBMs of latent image representations and text embeddings in each cross-attention layer of the denoising autoencoder. Our latent EBMs further allow zero-shot compositional generation as a linear combination of cross-attention outputs from different contexts.
  arXiv  Detail & Related papers  (2023-06-16T14:30:41Z)
• Diffusion Models for Open-Vocabulary Segmentation [79.02153797465324]
  OVDiff is a novel method that leverages generative text-to-image diffusion models for unsupervised open-vocabulary segmentation. It relies solely on pre-trained components and outputs the synthesised segmenter directly, without training.
  arXiv  Detail & Related papers  (2023-06-15T17:51:28Z)
• Task-Oriented Multi-Modal Mutual Leaning for Vision-Language Models [52.3032592038514]
  We propose a class-aware text prompt to enrich generated prompts with label-related image information. We achieve an average improvement of 4.03% on new classes and 3.19% on harmonic-mean over eleven classification benchmarks.
  arXiv  Detail & Related papers  (2023-03-30T06:02:40Z)
• Directed Diffusion: Direct Control of Object Placement through Attention Guidance [15.275386705641266]
  Text-guided diffusion models can generate an effectively endless variety of images given only a short text prompt describing the desired image content. These models often struggle to compose scenes containing several key objects such as characters in specified positional relationships. In this work, we take a particularly straightforward approach to providing the needed direction.
  arXiv  Detail & Related papers  (2023-02-25T20:48:15Z)
• Enhanced Modality Transition for Image Captioning [51.72997126838352]
  We build a Modality Transition Module (MTM) to transfer visual features into semantic representations before forwarding them to the language model. During the training phase, the modality transition network is optimised by the proposed modality loss. Experiments have been conducted on the MS-COCO dataset demonstrating the effectiveness of the proposed framework.
  arXiv  Detail & Related papers  (2021-02-23T07:20:12Z)

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.