RegionDrag is a region-based image editing technology jointly developed by the University of Hong Kong and the University of Oxford. Based on a diffusion model, it allows users to define handle and target regions to express their editing intentions, achieving fast and precise image editing...
LinFusion is an innovative image generation model developed by a research team at the National University of Singapore. It utilizes a linear attention mechanism to handle high-resolution image generation tasks, maintaining low computational complexity when processing large numbers of pixels...
Deepfake Defenders is an open-source AI model developed by the VisionRush team at the Institute of Automation, Chinese Academy of Sciences. It aims to identify and defend against forged images and videos generated by Deepfake technology. The model analyzes media...
The Blue Heart Big Model is a newly released self-developed general-purpose big model matrix from vivo, including language big models, device-side big models, speech big models, image big models, and multimodal big models. It plays a crucial role in multiple domains and scenarios...
VideoLLaMB is an innovative long-video understanding framework that processes video data by introducing a memory bridging layer and recursive memory tokens, ensuring that key visual information is not lost during analysis. The model is specifically designed for understanding long-form video...
MagicMan is an AI project jointly launched by research teams from Tsinghua University Shenzhen International Graduate School, Tencent AI Lab, Hong Kong University of Science and Technology, Stanford University, and the Chinese University of Hong Kong. It focuses on using deep learning technology to analyze single 2D images...
DeepSeek-Coder-V2 is an open-source code language model from DeepSeek, whose performance on code-specific tasks rivals that of GPT4-Turbo. The model is pre-trained on an additional 6 trillion...
AppFlowy is an open-source note-taking and task management tool designed as an alternative to Notion, offering better privacy. Built using Rust and Flutter, it adheres to minimalist principles and provides ample flexibility...
Fengyu Big Model is a large-scale language model specifically designed for the logistics industry, launched by SF Technology. SF Technology states that Fengyu Big Model, with its smaller size, comprehensively surpasses larger general-purpose models in the logistics vertical. Fengyu Big Model is already available in the market...
SAM2Point is a 3D segmentation technology based on SAM2. It performs zero-shot segmentation directly on arbitrary 3D data without additional training or 2D-3D projection. By voxelizing the 3D data and simulating a multi-directional video stream, it achieves precise segmentation based on SAM2...
CodeFuse-muAgent is a multi-agent framework developed by Ant Group's CodeFuse team. It uses a knowledge graph engine to drive the orchestration and collaboration of agents, simplifying the standard operating procedure orchestration process for agents. Through integration...
VoxInstruct is an open-source speech synthesis technology developed by Tsinghua University. It can generate highly user-relevant speech based on human language commands. The system employs a unified multilingual codec language modeling framework, transforming traditional text-to-speech...
Pixtral 12B is the first multimodal AI model launched by the French AI startup Mistral, capable of processing both images and text simultaneously. The model boasts 12 billion parameters, a size of approximately 24GB, and is built upon the text model Nemo 12B...
MMRole is a multimodal role-playing agent (MRPA) framework developed by the research team at the Gaoling School of Artificial Intelligence, Renmin University of China. By combining images and text, it enables agents to perform more natural and immersive roles...
RReHiFace-S is an open-source project launched by the Silicon Intelligence team that implements a high-fidelity, real-time face replacement AI algorithm, allowing users to swap faces in videos or live streams with simple operations.
NarratoAI is an AI-powered tool for automating video narration and editing. It uses a Large Language Model (LLM) to understand video content, automatically generates narration scripts, converts the scripts into voice-over, and simultaneously performs video editing and subtitling...
Skywork-Reward is a series of high-performance reward models launched by Kunlun Wanwei, including Skywork-Reward-Gemma-2-27B and Skywork-Reward-Llama-3.1-8B. It is primarily used to guide and optimize large language models...
Torch-MLU is an open-source PyTorch device backend extension plugin from Cambricon, allowing developers to use Cambricon's MLU series smart accelerator cards as PyTorch acceleration backends. The plugin implements native PyTorch support...
ImageBind is an open-source multimodal AI model from Meta that integrates six different types of information—text, audio, vision, temperature, and motion data—into a unified embedding space. The model uses image modalities as a bridge...
LlamaCoder is an open-source AI tool that uses the Llama 3.1 405B model to quickly generate full-stack applications. It aims to provide an alternative to Claude Artifacts. It integrates with Sandpack, Next...
XVERSE-MoE-A36B is China's largest open-source MoE (Mixture of Experts) model, launched by Yuanxiang. The model boasts 255 billion total parameters and 36 billion activation parameters, with performance comparable to models exceeding 100 billion...
PyVideoTrans is an open-source video translation and dubbing tool that automatically translates video content from one language to another and adds corresponding subtitles and dubbing. PyVideoTrans supports multiple languages and features efficient speech recognition...
ongCite is a project launched by Tsinghua University aimed at improving the credibility and verifiability of large language models (LLMs) in long-text question answering tasks. The project enables users to verify the model's performance by generating fine-grained sentence-level citations...
CSGO (Content-Style Composition in Text-to-Image Generation) is a research project on image style transfer and text-to-image generation, developed collaboratively by Nanjing University of Science and Technology, Xiaohongshu, and other institutions. CSGO's...