InstructCrop: Teaching Multimodal Large Language Models to Crop Aesthetic Images
Open Access
- 27 October 2025
- conference paper
- Published by Association for Computing Machinery (ACM)
- p. 6830-6839
Abstract
No abstract availableFunding Information
- National Natural Science Foundation of China (62471349, 62171340, and 62301378)
- Fundamental Research Funds for the Central Universities (QTZX25076)
- Guangzhou Municipal Government-University (Institute)-Enterprises Jointly Founded Project (2025A03J3123)
This publication has 45 references indexed in Scilit:
- Aesthetic image cropping meets VLP: Enhancing good while reducing badJournal of Visual Communication and Image Representation, 2024
- LISA: Reasoning Segmentation via Large Language Model2024
- Q-Instruct: Improving Low-Level Visual Abilities for Multi-Modality Foundation Models2024
- Composing Photos Like a Photographer2021
- Composing Good Shots by Exploiting Mutual Relations2020
- Reliable and Efficient Image Cropping: A Grid Anchor Based Approach2019
- Photographic composition classification and dominant geometric element detection for outdoor scenesJournal of Visual Communication and Image Representation, 2018
- A2-RL: Aesthetics Aware Reinforcement Learning for Image Cropping2018
- Deep Cropping via Attention Box Prediction and Aesthetics Assessment2017
- Convolutional Neural Networks for No-Reference Image Quality Assessment2014