BIT unifies text-to-image and image-to-text in one bidirectional diffusion
burny_tech · x · 2026-09-01
New research introduces BIT (Bidirectional Image-Text Diffusion Bridges), unifying text-to-image and image-to-text generation. Instead of starting from noise, the model transitions directly between text tokens and images within a single bidirectional diffusion process. Paper, demo, and code are available.
Related event: BIT Framework Unifies Text-to-Image and Image-to-Text Generation(2 posts)→
More from Multimodal
- Complex multi-link CAD mechanism finally animates, dev celebrates the breakthrough — zetalyrae · 2026-10-16
- OpenMOSS releases MOSS-Image-SFT: Apache 2.0 text-to-image model on Hugging Face — bdsqlsz · 2026-10-16
- Hyperflow open-sourced: 49 sampling steps down to 8, ~3x faster video generation — mhdfaran · 2026-10-16
- Cinematic 3D fantasy character prompt template with salvaged-robot armor — azed_ai · 2026-10-16
- Midjourney vintage woodcut style: --sref 2446102396 beaver with a cleaver — michaelrabone · 2026-10-16
- MOSS-Image INT8 quantization lands for ComfyUI, near-lossless per author — East_Shoe8814 · 2026-10-16