Nano Banana keeps failing face identity: a stylized 3D face-swap pipeline question
LocksmithOk4723 · reddit · 2026-11-17
A developer building an AI image pipeline reports that Google's Nano Banana models (Flash/Pro) preserve scene composition, objects, and pose well but consistently fail to keep the reference photo's facial identity — even with mask-based editing restricted to face and hair. Goal: consistent identity across multiple stylized 3D scenes from a single photo. Candidate directions listed: InstantID, IP-Adapter FaceID, PuLID, and ControlNet + face embeddings + inpainting combos, via ComfyUI or a custom pipeline.
More from Multimodal
- After DALL-E Shutdown Forced a Pipeline Rewrite: Four Questions for Picking Generative Models — Al_Grigor · 2026-11-17
- 80s-style fantasy AI video dubbed better than movies — mark_k · 2026-11-17
- New audio encoders called a huge leap: embeddings far more informative, audio-CoCa teased — wightmanr · 2026-11-17
- The Burrito Test: one prompt stress-tests 25 image generation models side by side — generativist · 2026-11-17
- A WW1 rap video made entirely with local ComfyUI models, free of charge — Motor_Mix2389 · 2026-11-17
- AI-generated Link selfie in Hyrule fools Reddit with uncanny Zelda cosplay video — Mattapeh · 2026-11-17