전체 글(24)
-
2D image to 3D model (mesh) AI free : Point E
Point E : Automatically convert 2D image to 3D object modeling mesh with AI. Powered by machine-learning algorithms, this enables developers to create their 3D modeling mesh in a quick and free. This image-to-3D model was fed a set of images paired with 3D objects mesh so that it learned to effectively translate. GitHub - openai/point-e: Point cloud diffusion for 3D model synthesis GitHub - open..
2022.12.21 -
Diffusers v0.11.0 Released
Diffusers v0.11.0 Releases * Karlo UnCLIP, safetensors, pipeline versions Releases · huggingface/diffusers (github.com) GitHub - huggingface/diffusers: 🤗 Diffusers: State-of-the-art diffusion models for image and audio generation in PyTorch 🤗 Diffusers: State-of-the-art diffusion models for image and audio generation in PyTorch - GitHub - huggingface/diffusers: 🤗 Diffusers: State-of-the-art diff..
2022.12.20 -
How to run TensorFlow on Apple Mac M1, M2 with GPU support
How to run TensorFlow on Apple Mac M1, M2 with GPU support 1. Xcode command-line tools: xcode-select --install 2. anaconda(miniconda) setup https://repo.anaconda.com/miniconda/Miniconda3-latest-MacOSX-arm64.sh bash ~/miniconda.sh -b -p $HOME/miniconda source ~/miniconda/bin/activate 3. Install tensorflow-deps for apple mac conda install -c apple tensorflow-deps 4. Install TensorFlow (Base) pytho..
2022.12.15 -
InvokeAI 2.2.4 (2022-12-12) Release
InvokeAI is a fork of CompVis/stable-diffusion, the open source text-to-image generator. It provides a streamlined process with various new features and options to aid the image generation process. It runs on Windows, Apple MacOS and Linux. It provides both a polished web interface(Web-UI), and a command-line interface. Getting Started with InvokeAI 1. Windows 1-1. Download for Windows https://g..
2022.12.13 -
hand style : dall-e2 vs stable diffusion
hand style : dall-e2 vs stable diffusion Dall-e2가 스탁 이미지 등 명확한 표현을 보장하는 데이터셋이 4배 정도 더 많고, Stable Diffusion의 Default CFG가 손 모양 표현에 적합하지 않기 때문에, Stable Diffusion에서 손 모양을 개선하는 방법은 손에 대한 표현이 풍부한 라벨링 데이터, 손 모양에 대한 스타일 데이터셋의 충분량 필요, 그리고 손이나 발 모양 표현에 적합한 Steps와 CFG로 설정하는 것이 좋다. 다음은 Stable Diffusion 2.1에서 많은 개선이 이뤄졌다는 손 모양에 대한 여러가지 출력결과(개선된 로직이 동일 데이터셋 문제로 인하여 개선의 정도를 알 수 없을 정도..) Stable Diffusion 2.1 p..
2022.12.11