Hugging Face's logo
Nikita Mikhaylov's picture
13 181

Nikita Mikhaylov

mihaylovnikitos
shtefcs's profile picture 21world's profile picture
ยท

AI & ML interests

None yet

Recent Activity

liked a model 2 days ago
microsoft/phi-4
reacted to sanaka87's post with ๐Ÿš€ 3 days ago
๐Ÿš€ Excited to Share Our Latest Work: 3DIS & 3DIS-FLUX for Multi-Instance Layout-to-Image Generation! โค๏ธโค๏ธโค๏ธ ๐ŸŽจ Daily Paper: https://huggingface.co/papers/2501.05131#community ๐Ÿ”“ Code is now open source! ๐ŸŒ Project Website: https://limuloo.github.io/3DIS/ ๐Ÿ  GitHub Repository: https://github.com/limuloo/3DIS ๐Ÿ“„ 3DIS Paper: https://arxiv.org/abs/2410.12669 ๐Ÿ“„ 3DIS-FLUX Tech Report: https://arxiv.org/abs/2501.05131 ๐Ÿ”ฅ Why 3DIS & 3DIS-FLUX? Current SOTA multi-instance generation methods are typically adapter-based, requiring additional control modules trained on pre-trained models for layout and instance attribute control. However, with the emergence of more powerful models like FLUX and SD3.5, these methods demand constant retraining and extensive resources. โœจ Our Solution: 3DIS We introduce a decoupled approach that only requires training a low-resolution Layout-to-Depth model to convert layouts into coarse-grained scene depth maps. Leveraging community and company pre-trained models like ControlNet + SAM2, we enable training-free controllable image generation on high-resolution models such as SDXL and FLUX. ๐ŸŒŸ Benefits of Our Decoupled Multi-Instance Generation: 1. Enhanced Control: By constructing scenes using depth maps in the first stage, the model focuses on coarse-grained scene layout, improving control over instance placement. 2. Flexibility & Preservation: The second stage employs training-free rendering methods, allowing seamless integration with various models (e.g., fine-tuned weights, LoRA) while maintaining the generative capabilities of pre-trained models. Join us in advancing Layout-to-Image Generation! Follow and star our repository to stay updated! โญ
reacted to sanaka87's post with ๐Ÿ”ฅ 3 days ago
๐Ÿš€ Excited to Share Our Latest Work: 3DIS & 3DIS-FLUX for Multi-Instance Layout-to-Image Generation! โค๏ธโค๏ธโค๏ธ ๐ŸŽจ Daily Paper: https://huggingface.co/papers/2501.05131#community ๐Ÿ”“ Code is now open source! ๐ŸŒ Project Website: https://limuloo.github.io/3DIS/ ๐Ÿ  GitHub Repository: https://github.com/limuloo/3DIS ๐Ÿ“„ 3DIS Paper: https://arxiv.org/abs/2410.12669 ๐Ÿ“„ 3DIS-FLUX Tech Report: https://arxiv.org/abs/2501.05131 ๐Ÿ”ฅ Why 3DIS & 3DIS-FLUX? Current SOTA multi-instance generation methods are typically adapter-based, requiring additional control modules trained on pre-trained models for layout and instance attribute control. However, with the emergence of more powerful models like FLUX and SD3.5, these methods demand constant retraining and extensive resources. โœจ Our Solution: 3DIS We introduce a decoupled approach that only requires training a low-resolution Layout-to-Depth model to convert layouts into coarse-grained scene depth maps. Leveraging community and company pre-trained models like ControlNet + SAM2, we enable training-free controllable image generation on high-resolution models such as SDXL and FLUX. ๐ŸŒŸ Benefits of Our Decoupled Multi-Instance Generation: 1. Enhanced Control: By constructing scenes using depth maps in the first stage, the model focuses on coarse-grained scene layout, improving control over instance placement. 2. Flexibility & Preservation: The second stage employs training-free rendering methods, allowing seamless integration with various models (e.g., fine-tuned weights, LoRA) while maintaining the generative capabilities of pre-trained models. Join us in advancing Layout-to-Image Generation! Follow and star our repository to stay updated! โญ
View all activity

Organizations

None yet

mihaylovnikitos's activity

upvoted a collection 3 months ago
upvoted an article 5 months ago
upvoted 2 collections 6 months ago
upvoted 3 articles 7 months ago
view article
Article

How I train a LoRA: m3lt style training overview

By alvdansen โ€ข
โ€ข 49
view article
Article

Recommendation to Revisit the Diffuser Default LoRA Parameters

By alvdansen โ€ข
โ€ข 11
view article
Article

Putting RL back in RLHF

โ€ข 69
upvoted an article 8 months ago
view article
Article

Training and Finetuning Embedding Models with Sentence Transformers v3

โ€ข 173
upvoted 3 articles 9 months ago
view article
Article

Outpainting I - Controlnet version

By OzzyGT โ€ข
โ€ข 13
upvoted a paper 9 months ago
upvoted a collection 10 months ago