Just 2,305 new params: open-source LoRA teaches FLUX.2 Klein to cut objects into transparent PNGs
ostrisai · x · 2026-09-30
Developer trmz shared a neat open-source project: teaching FLUX.2 Klein to see transparency and extract foregrounds.
- The model's VAE was widened from RGB to RGBA, adding only 2,305 new parameters out of 84M
- A LoRA trained on FLUX.2 Klein 9B cuts objects out of photos as real transparent PNGs, shadows included
- Fully open source
More from Multimodal
- GPT-6.1 Sol near-Astra quality in 3D work while running ~30% faster — cedric_chee · 2026-09-30
- Midjourney + Kling 4.0 Flash combo wows creators in video generation tests — gen_ericai · 2026-09-30
- UniMate open-sources unified text-to-animation model that drives diverse 3D skeletons — grandorganics · 2026-09-30
- Redditor creates original Spanish-language AI animated film with ComfyUI, Flux and MiniMax H3 — Low_Masterpiece_7861 · 2026-09-30
- fal opens its Agent to everyone: image, video, audio and 3D in one interface — adamho · 2026-09-30
- AssemblyAI launches Universal-3.6 Pro Realtime, tops two streaming speech benchmarks — AssemblyAI · 2026-09-30