Progressive Prompt Detailing for Improved Alignment in Text-to-Image Generative Models

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Saichandran, Ketan Suhaas, Thomas, Xavier, Kaushik, Prakhar, Ghadiyaram, Deepti
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866918039754113024
author Saichandran, Ketan Suhaas
Thomas, Xavier
Kaushik, Prakhar
Ghadiyaram, Deepti
author_facet Saichandran, Ketan Suhaas
Thomas, Xavier
Kaushik, Prakhar
Ghadiyaram, Deepti
contents Text-to-image generative models often struggle with long prompts detailing complex scenes, diverse objects with distinct visual characteristics and spatial relationships. In this work, we propose SCoPE (Scheduled interpolation of Coarse-to-fine Prompt Embeddings), a training-free method to improve text-to-image alignment by progressively refining the input prompt in a coarse-to-fine-grained manner. Given a detailed input prompt, we first decompose it into multiple sub-prompts which evolve from describing broad scene layout to highly intricate details. During inference, we interpolate between these sub-prompts and thus progressively introduce finer-grained details into the generated image. Our training-free plug-and-play approach significantly enhances prompt alignment, achieves an average improvement of more than +8 in Visual Question Answering (VQA) scores over the Stable Diffusion baselines on 83% of the prompts from the GenAI-Bench dataset.
format Preprint
id arxiv_https___arxiv_org_abs_2503_17794
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Progressive Prompt Detailing for Improved Alignment in Text-to-Image Generative Models
Saichandran, Ketan Suhaas
Thomas, Xavier
Kaushik, Prakhar
Ghadiyaram, Deepti
Computer Vision and Pattern Recognition
Artificial Intelligence
Text-to-image generative models often struggle with long prompts detailing complex scenes, diverse objects with distinct visual characteristics and spatial relationships. In this work, we propose SCoPE (Scheduled interpolation of Coarse-to-fine Prompt Embeddings), a training-free method to improve text-to-image alignment by progressively refining the input prompt in a coarse-to-fine-grained manner. Given a detailed input prompt, we first decompose it into multiple sub-prompts which evolve from describing broad scene layout to highly intricate details. During inference, we interpolate between these sub-prompts and thus progressively introduce finer-grained details into the generated image. Our training-free plug-and-play approach significantly enhances prompt alignment, achieves an average improvement of more than +8 in Visual Question Answering (VQA) scores over the Stable Diffusion baselines on 83% of the prompts from the GenAI-Bench dataset.
title Progressive Prompt Detailing for Improved Alignment in Text-to-Image Generative Models
topic Computer Vision and Pattern Recognition
Artificial Intelligence
url https://arxiv.org/abs/2503.17794