Vision as Unified Multimodal Generation
We formulate computer vision as unified multimodal generation, where heterogeneous visual tasks are expressed in the native text and image generation spaces of a unified multimodal model, without task-specific architectures. Under this formulation, SenseNova-Vision uses natural-language instructions and optional visual prompts to specify tasks, target regions or views, and decoding conventions, and generates responses as text for symbolic outputs, images for dense spatial predictions, or mixed text-and-image outputs for compositional tasks. To support large-scale training, we convert diverse c
Lineage graph
Paper → model → repo connections mined from source citations (Tier-1 exact match).
Why these links exist
- Linked via arxiv authorXiaoyang Han →
Vision as Unified Multimodal Generation
- Linked via arxiv authorJianhua Li →
Vision as Unified Multimodal Generation
- Linked via arxiv authorKewang Deng →
Vision as Unified Multimodal Generation
- Linked via arxiv authorZukai Chen →
Vision as Unified Multimodal Generation
- Linked via arxiv authorXuanke Shi →
Vision as Unified Multimodal Generation
- Linked via arxiv authorSihan Wang →
Vision as Unified Multimodal Generation
- Linked via arxiv authorBoxuan Li →
Vision as Unified Multimodal Generation
- Linked via arxiv authorLinyan Wang →
Vision as Unified Multimodal Generation
- Linked via arxiv authorSiyi Xie →
Vision as Unified Multimodal Generation
- Linked via arxiv authorXin You →
Vision as Unified Multimodal Generation
- Linked via arxiv authorJinsheng Quan →
Vision as Unified Multimodal Generation
- Linked via arxiv authorZhongang Cai →
Vision as Unified Multimodal Generation
- Linked via arxiv authorHaiwen Diao →
Vision as Unified Multimodal Generation
- Linked via arxiv authorZiwei Liu →
Vision as Unified Multimodal Generation
- Linked via arxiv authorLei Yang →
Vision as Unified Multimodal Generation
- Linked via arxiv authorDahua Lin →
Vision as Unified Multimodal Generation
- Linked via arxiv authorQuan Wang →
Vision as Unified Multimodal Generation