Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2504.08003 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866915236773101568 |
|---|---|
| author | Li, Ning Zhang, Jingran Cui, Justin |
| author_facet | Li, Ning Zhang, Jingran Cui, Justin |
| contents | OpenAI's multimodal GPT-4o has demonstrated remarkable capabilities in image generation and editing, yet its ability to achieve world knowledge-informed semantic synthesis--seamlessly integrating domain knowledge, contextual reasoning, and instruction adherence--remains unproven. In this study, we systematically evaluate these capabilities across three critical dimensions: (1) Global Instruction Adherence, (2) Fine-Grained Editing Precision, and (3) Post-Generation Reasoning. While existing benchmarks highlight GPT-4o's strong capabilities in image generation and editing, our evaluation reveals GPT-4o's persistent limitations: the model frequently defaults to literal interpretations of instructions, inconsistently applies knowledge constraints, and struggles with conditional reasoning tasks. These findings challenge prevailing assumptions about GPT-4o's unified understanding and generation capabilities, exposing significant gaps in its dynamic knowledge integration. Our study calls for the development of more robust benchmarks and training strategies that go beyond surface-level alignment, emphasizing context-aware and reasoning-grounded multimodal generation. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2504_08003 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Have we unified image generation and understanding yet? An empirical study of GPT-4o's image generation ability Li, Ning Zhang, Jingran Cui, Justin Computer Vision and Pattern Recognition OpenAI's multimodal GPT-4o has demonstrated remarkable capabilities in image generation and editing, yet its ability to achieve world knowledge-informed semantic synthesis--seamlessly integrating domain knowledge, contextual reasoning, and instruction adherence--remains unproven. In this study, we systematically evaluate these capabilities across three critical dimensions: (1) Global Instruction Adherence, (2) Fine-Grained Editing Precision, and (3) Post-Generation Reasoning. While existing benchmarks highlight GPT-4o's strong capabilities in image generation and editing, our evaluation reveals GPT-4o's persistent limitations: the model frequently defaults to literal interpretations of instructions, inconsistently applies knowledge constraints, and struggles with conditional reasoning tasks. These findings challenge prevailing assumptions about GPT-4o's unified understanding and generation capabilities, exposing significant gaps in its dynamic knowledge integration. Our study calls for the development of more robust benchmarks and training strategies that go beyond surface-level alignment, emphasizing context-aware and reasoning-grounded multimodal generation. |
| title | Have we unified image generation and understanding yet? An empirical study of GPT-4o's image generation ability |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2504.08003 |