ProPainter is an AI video restoration project developed by the S-Lab team at Nanyang Technological University. By combining dual-domain propagation and a mask-guided sparse video Transformer, it effectively improves the performance of video restoration. ProPainter can automatically...
auto-video-generator is an AI-powered automatic narration video generator that can generate narration videos with a single click. Users simply input a topic, and the system automatically writes the script, synthesizes the speech, generates images, and merges them into a video, greatly improving the efficiency of video creation...
Roop is an open-source AI video face-swapping tool. Roop allows users to replace faces in videos with a single image, without requiring complex datasets or training processes. Installation and use require some technical skills, making it suitable for users with some basic knowledge...
VFusion3D is an AI-generated 3D model project launched by researchers from Meta and Oxford University. It can generate high-quality 3D objects from single images or text descriptions. VFusion3D fine-tunes pre-trained video AI...
TurboEdit is an AI-powered real-time image editing model developed by Adobe Research. Through iterative encoder inversion and fine-grained text-based control, it enables precise image editing in just a few steps. Utilizing detailed text prompts,...
edge-tts is an open-source AI text-to-speech project that supports over 40 languages and more than 300 voices. Edge-tts leverages the power of Microsoft Azure Cognitive Services to convert text information into fluent and natural speech...
Wav2Lip is an open-source lip-syncing tool that allows users to convert audio files into videos synchronized with lip movements. It is widely used in video editing and game development. Wav2Lip not only enables real-time lip-syncing but also supports...
LM Studio is a native large language model (LLM) application platform that is open-source, user-friendly, and offers a one-stop solution for deploying native large models. It supports, but is not limited to, Llama, MPT, Gemma, etc. LM Studio provides a graphical user interface...
AuraFlow v0.1 is an open-source AI text-to-image generation model from the Fal team, boasting 6.8B parameters. It optimizes the MMDiT architecture, improving computational efficiency and scalability. AuraFlow excels at accurate image generation, especially in object...
LongWriter is a long text generation model jointly developed by Tsinghua University and Zhipu AI. It can generate coherent text exceeding 10,000 characters, and the project is open source. By analyzing the output limitations of existing large-scale language models, 'LongWriter-...' was created.
Source 2.0-M32 is a hybrid expert model (MoE) launched by Inspur Information, featuring 32 experts. It employs an innovative 'Attention Router' technology to improve the efficiency and accuracy of expert selection. The model has a total of 4 billion parameters...
Zion is a no-code AI application development platform that significantly lowers the technical barrier to application development through visual configuration and drag-and-drop interaction. Users can quickly build mini-programs, websites, and other applications without any programming knowledge. Zion integrates...
Dify is an open-source generative AI application development platform that helps users quickly build and deploy applications based on large language models. It supports multiple mainstream models, such as DeepSeek and Ollama, allowing users to flexibly choose according to their needs...
Video-LLaVA2 is an open-source multimodal intelligent understanding system developed by the ChatLaw research group at Peking University. It enhances video and audio understanding capabilities through innovative spatiotemporal convolution (STC) connectors and audio branches. The model performs well in video...
MUMU is a multimodal image generation model that improves the accuracy and quality of generated images by combining text prompts and reference images. The MUMU model architecture is based on the pre-trained convolutional UNet from SDXL and employs...
LLaVA-OneVision is an open-source multimodal AI model launched by ByteDance. By integrating insights from data, models, and visual representations, LLaVA-OneVision can simultaneously handle computer vision scenarios involving single images, multiple images, and video...
Insee AIGC is an AI marketing tool launched by Insee Group. InsightGPT combines AI technology with marketing strategy insights, content creation and production, dissemination and delivery, and performance conversion to improve the productivity of the marketing industry. InsightGPT not only has...
ComfyUI-Bxb is an "SD monetization tool" that converts ComfyUI workflows into mini-programs with a single click. Designed specifically for mini-program developers, it enables rapid monetization. The core function of ComfyUI-Bxb is to convert ComfyUI workflows into WeChat mini-programs...
HMoE (Hybrid Heterogeneous Expert Model) is a novel neural network architecture proposed by Tencent's Hunyuan team, aiming to improve the performance and computational efficiency of large-scale language models. It introduces experts of different sizes to handle inputs of varying complexity...
Transfusion is Meta's latest multimodal AI model for text and image fusion. Transfusion processes mixed-modal data on a single transformer by combining next-to-toe prediction from a language model with a diffusion model...
RAGFlow is an open-source RAG (Retrieval-Augmented Generation) engine with deep document understanding and enhanced generation capabilities. It is a next-generation RAG engine based on OCR and document parsing. RAGFlow integrates its own...
StyleShot is an open-source AI image style transfer model that can transfer any style to any content without additional training. It extracts style features through a style-aware encoder and enhances the style-content fusion encoder...
IMAGDressing is an AI-powered clothing-changing tool launched by Huawei in partnership with Tencent. IMAGDressing allows users to generate clothing in a modular fashion and try it on in a virtual environment. The project page provides automated workflows, code examples, and more.
NextChat is an open-source project based on Next.js and Vercel, allowing users to integrate large AI models such as ChatGPT into their web applications. NextChat not only provides a ready-made codebase but also supports one-click integration...