{"slug": "dataset-lora-training-pipeline-for-room-emptying-object-removal-image-to-image", "title": "Dataset & LoRA Training Pipeline for Room Emptying / Object Removal (Image-to-Image)", "summary": "A community member is seeking guidance on dataset sources and training architecture for a project that empties indoor spaces and removes specific objects from images using provided object masks. The poster asks whether paired furnished-room/empty-room datasets such as InteriorNet, Structured3D, or Matterport3D suit an image-to-image or diffusion pipeline, and whether fine-tuning a LoRA on an inpainting model like SDXL Inpainting or Flux Inpainting with mask-guided conditioning is preferable to training a custom ControlNet. The request also asks for relevant papers, datasets, and codebases.", "body_md": "**Hi Community,**\n\nI am working on a project aimed at **emptying indoor spaces** (removing all furniture/decor) as well as **removing specific individual objects** from an image using provided object masks.\n\nI plan to fine-tune a model (or train a LoRA / ControlNet adapter), but I need guidance on dataset sources and the ideal training architecture.\n\nHere are my specific questions:\n\n**Dataset Sources for Empty Room Pairs:**\n\nWhere can I find datasets containing paired images of **[Furnished Room ↔ Empty Room]** suitable for training an image-to-image or diffusion pipeline?\n\nAre there existing synthetic or 3D-rendered datasets (e.g., InteriorNet, Structured3D, Matterport3D) recommended for this specific paired generation task?\n\n**Single Object Removal & Inpainting:**\n\nFor the use case of removing a single specific object where I already have the **object mask**, what is the recommended architecture?\n\nIs fine-tuning a LoRA on an Inpainting model (like SDXL Inpainting or Flux Inpainting) using mask-guided conditioning preferable over training a custom ControlNet?\n\nAny pointers to relevant papers, datasets, or codebases would be greatly appreciated!\n\nThanks!", "url": "https://wpnews.pro/news/dataset-lora-training-pipeline-for-room-emptying-object-removal-image-to-image", "canonical_source": "https://discuss.huggingface.co/t/dataset-lora-training-pipeline-for-room-emptying-object-removal-image-to-image/180659#post_1", "published_at": "2026-09-21 06:51:09+00:00", "updated_at": "2026-09-21 06:52:36.425313+00:00", "lang": "en", "topics": ["computer-vision", "generative-ai", "ai-research", "ai-tools"], "entities": ["InteriorNet", "Structured3D", "Matterport3D", "SDXL Inpainting", "Flux Inpainting", "ControlNet", "LoRA"], "alternates": {"html": "https://wpnews.pro/news/dataset-lora-training-pipeline-for-room-emptying-object-removal-image-to-image", "markdown": "https://wpnews.pro/news/dataset-lora-training-pipeline-for-room-emptying-object-removal-image-to-image.md", "text": "https://wpnews.pro/news/dataset-lora-training-pipeline-for-room-emptying-object-removal-image-to-image.txt", "jsonld": "https://wpnews.pro/news/dataset-lora-training-pipeline-for-room-emptying-object-removal-image-to-image.jsonld"}}