Text-Image Conditioned 3D Generation
Fuente:
arXiv
Salvato in:
| Autori principali: | , , , , , , , , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2026
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866910062955462656 |
|---|---|
| author | Cen, Jiazhong Fang, Jiemin Li, Sikuang Wu, Guanjun Yang, Chen Yi, Taoran Zhou, Zanwei Bao, Zhikuan Xie, Lingxi Shen, Wei Tian, Qi |
| author_facet | Cen, Jiazhong Fang, Jiemin Li, Sikuang Wu, Guanjun Yang, Chen Yi, Taoran Zhou, Zanwei Bao, Zhikuan Xie, Lingxi Shen, Wei Tian, Qi |
| contents | High-quality 3D assets are essential for VR/AR, industrial design, and entertainment, motivating growing interest in generative models that create 3D content from user prompts. Most existing 3D generators, however, rely on a single conditioning modality: image-conditioned models achieve high visual fidelity by exploiting pixel-aligned cues but suffer from viewpoint bias when the input view is limited or ambiguous, while text-conditioned models provide broad semantic guidance yet lack low-level visual detail. This limits how users can express intent and raises a natural question: can these two modalities be combined for more flexible and faithful 3D generation? Our diagnostic study shows that even simple late fusion of text- and image-conditioned predictions outperforms single-modality models, revealing strong cross-modal complementarity. We therefore formalize Text-Image Conditioned 3D Generation, which requires joint reasoning over a visual exemplar and a textual specification. To address this task, we introduce TIGON, a minimalist dual-branch baseline with separate image- and text-conditioned backbones and lightweight cross-modal fusion. Extensive experiments show that text-image conditioning consistently improves over single-modality methods, highlighting complementary vision-language guidance as a promising direction for future 3D generation research. Project page: https://jumpat.github.io/tigon-page |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2603_21295 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Text-Image Conditioned 3D Generation Cen, Jiazhong Fang, Jiemin Li, Sikuang Wu, Guanjun Yang, Chen Yi, Taoran Zhou, Zanwei Bao, Zhikuan Xie, Lingxi Shen, Wei Tian, Qi Computer Vision and Pattern Recognition High-quality 3D assets are essential for VR/AR, industrial design, and entertainment, motivating growing interest in generative models that create 3D content from user prompts. Most existing 3D generators, however, rely on a single conditioning modality: image-conditioned models achieve high visual fidelity by exploiting pixel-aligned cues but suffer from viewpoint bias when the input view is limited or ambiguous, while text-conditioned models provide broad semantic guidance yet lack low-level visual detail. This limits how users can express intent and raises a natural question: can these two modalities be combined for more flexible and faithful 3D generation? Our diagnostic study shows that even simple late fusion of text- and image-conditioned predictions outperforms single-modality models, revealing strong cross-modal complementarity. We therefore formalize Text-Image Conditioned 3D Generation, which requires joint reasoning over a visual exemplar and a textual specification. To address this task, we introduce TIGON, a minimalist dual-branch baseline with separate image- and text-conditioned backbones and lightweight cross-modal fusion. Extensive experiments show that text-image conditioning consistently improves over single-modality methods, highlighting complementary vision-language guidance as a promising direction for future 3D generation research. Project page: https://jumpat.github.io/tigon-page |
| title | Text-Image Conditioned 3D Generation |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2603.21295 |