CivArchive
    ← All articles
    Published September 6, 2024by DMphotoart

    News: ,

    380 views2 reactions0 comments on CivitAI1 collected
    newslistresource searchnews,flux ai

    News:

    • SKYBOX AI: Create 360° worlds from a single image. Try it here

    • Text-Guided-Image-Colorization: Colorize objects in your images using text prompts (SDXL + CLIP). GitHub

    • Meta’s Sapiens Segmentation Model: Available now on Hugging Face Spaces. [Demo](HUGGING FACE DEMO)

    • Anifusion.ai: Create comic books via web app. Explore it here

    • MiniMax: New Chinese text-to-video model + free music generation. Video Model | Music Model

    • Viewcrafter: Generate high-fidelity views from sparse input images with camera control. [GitHub Code](GITHUB CODE) | [Hugging Face Demo](HUGGING FACE DEMO)

    • LumaLabsAI Dream Machine V6.1: Now featuring camera controls.

    • RB-Modulation by Google: Training-free diffusion model personalization using stochastic optimal control. [Hugging Face Demo](HUGGING FACE DEMO)

    • New ChatGPT Voices: Fathom, Glimmer, Harp, Maple, Orbit, Rainbow (1, 2, 3), Reef, Ridge, and Vale. (X Video Preview)

    • FluxMusic: State-of-the-art open-source text-to-music model. GitHub | [Jupyter Notebook](JUPYTER NOTEBOOK) | Paper

    • P2P-Bridge: Denoise 3D scans. GitHub | Paper

    • HivisionIDPhoto: Model set for portrait recognition, image cutout, and ID photo generation. [Hugging Face Demo](HUGGING FACE DEMO) | GitHub

    • ComfyUI-AdvancedLivePortrait Update: GitHub

    • ComfyUI v0.2.0: Adds Flux controlnets, queue management improvements, and node library enhancements. [Blog Post](BLOG POST)

    • SUNO: Their AI-generated song hit 100k views on YouTube. Watch here

    Catch all these updates in this week’s newsletter. Check out the latest issue for more!

    Updates from Last Week:

    • Joy Caption Update: Faster, improved natural language captions for images, including NSFW content, with ComfyUI integration.

    • FLUX Training Insights: New findings suggest FLUX understands complex concepts better than anticipated.

    • Realism Techniques: Tips on generating more realistic images with FLUX by reducing guidance scale and image quality in prompts.

    • LoRA Training for Logos: Discussion on using FLUX for training company logos, including dataset size and parameters.

    Additional Tools & Models:

    • FluxForge v0.1: A tool for searching FLUX LoRA models across Civitai and Hugging Face, updated every 2 hours.

    • Juggernaut XI: Enhanced SDXL model with improved prompt adherence.

    • FLUX.1 AI-Toolkit UI on Gradio: Drag-and-drop functionality and AI captioning.

    • Kolors Virtual Try-On App UI on Gradio: Virtual clothing try-on demo.

    • CogVideoX-5B: Open-weights text-to-video model generating 6-second videos.

    • Melyn’s 3D Render SDXL LoRA: LoRA model for Stable Diffusion XL, trained on personal 3D renders.

    • sd-ppp Photoshop Extension: Adds regional prompt support for ComfyUI in Photoshop.

    • GenWarp: Generates new scene viewpoints from a single input image.

    • Flux Latent Detailer Workflow: Experimental ComfyUI workflow for enhancing image details with latent interpolation.