iDP3 (Improved 3D Diffusion Policy) is a 3D vision-motor strategy developed by Stanford University in collaboration with several other universities. It enhances the autonomous operation capabilities of humanoid robots in diverse environments. Unlike traditional 3D strategies, iDP3...
Proactive Agent is a next-generation proactive agent interaction paradigm developed by Tsinghua University in collaboration with teams such as Facewall Intelligence. It possesses initiative, can predict user needs, and take action without direct commands. Proactive Agent...
EMOVA (Emotionally Omni-present Voice Assistant) is a multimodal, all-around model jointly developed by the Hong Kong University of Science and Technology, the University of Hong Kong, and Huawei Noah's Ark Lab, among other institutions. EMOVA can process images,...
OminiControl is a highly versatile and parameter-efficient image generation framework designed for diffusion transformer models such as FLUX.1, enabling fine-grained control over the image generation process. OminiControl supports subject-driven control and spatial control...
Talker-Reasoner is an AI agent architecture developed by Google DeepMind. Drawing inspiration from human cognitive theories, it divides the agent into two modules: Talker and Reasoner. Talker simulates rapid, intuitive human thinking (System 1),...
Diffusion Self-Distillation (DSD) is an innovative zero-shot customized image generation technique that automatically generates datasets using a pre-trained text-to-image diffusion model and fine-tunes them to enable text-conditional image-to-image processing...
Open Materials 2024 (OMat24) is a large open dataset released by MetaData containing density functional theory (DFT) calculations for over 110 million structures, focusing on the structural and compositional diversity of inorganic materials. It includes pre-trained...
CodeDPO is a code generation and optimization framework developed in collaboration between Peking University and ByteDance. It improves the correctness and efficiency of code models. The framework is based on a self-generation and verification mechanism, simultaneously building and evaluating code and its tests...
Voice-Pro is an open-source, multi-functional audio processing tool that integrates speech-to-text (STT), text-to-speech (TTS), real-time translation, YouTube video downloading, and voice separation, among other features. The tool supports over 100 languages...
Generative Omnimatte is a video editing technology developed by Google DeepMind and other organizations. It can intelligently decompose a video into multiple RGBA layers with transparent backgrounds, each layer corresponding to an object and its related effects (such as shadows, shadows, etc.).
GLM-PC is a computer intelligence agent developed by Zhipu, based on the multimodal large-scale model CogAgent. It can "observe" and "operate" computers like a human, assisting users in efficiently completing various computer tasks, such as document processing, web searching, information processing, etc.
TryOffDiff (VTOFF) is a novel virtual try-on technology based on a diffusion model. It uses high-fidelity clothing reconstruction to achieve virtual try-on, focusing on generating standardized clothing images from a single photograph of the wearer. Compared to traditional Virtual Try-Off...
GLM-Edge is a series of edge-deployment optimized large language dialogue models and multimodal understanding models from Zhipu Open Source, including GLM-Edge-1.5B-Chat, GLM-Edge-4B-Chat, GLM-Edge-V-2B, and GLM-Edge-V-5B. 1.5...
FlagEvalMM is an open-source multimodal model evaluation framework from the Beijing Academy of Artificial Intelligence. It comprehensively evaluates models handling multiple modalities, including text, images, and video, and supports various tasks and metrics. The framework employs a combination of evaluation and model...
Delta-CoMe is a novel incremental compression algorithm proposed by the Tsinghua University NLP Lab in collaboration with the OpenBMB open-source community, Peking University, and Shanghai University of Finance and Economics. An 80GB A100 GPU can easily load up to 50 7B models, saving GPU memory...
Agent-E is an intelligent automation system built on the AutoGen proxy framework, focusing on automated operations within the browser. Based on natural language interaction, Agent-E can perform tasks such as filling out forms, searching and ranking e-commerce products, and locating webpages...
Magic Copy is an open-source image cutout tool that supports Chrome browser extensions. Based on Meta's Segment Anything Model technology, it automatically identifies and extracts foreground objects from images and copies them to the user's clipboard...
NSFW Detector (Not Safe For Work) is an open-source tool for detecting inappropriate content. NSFW Detector can identify inappropriate content in images, PDFs, and video files. The tool is based on Google's Vita-...
MultiFoley is a sound generation system jointly developed by Adobe Research and the University of Michigan. It can generate Foley sound effects based on multimodal control of text, audio, and video. The system allows users to generate Foley sound effects based on text prompts, reference audio, etc.
CAT4D, developed jointly by Google DeepMind, Columbia University, and the University of California, San Diego, is a technology that creates 4D scene (dynamic 3D) representations from monocular video. CAT4D is based on a multi-view video diffusion model, enabling the creation of 4D scene representations (dynamic 3D) from any point of view...
Make-It-Animatable is a data-driven framework developed by the University of Science and Technology of China and Tencent PCG. It can prepare any 3D humanoid model for character animation in less than a second, regardless of its shape and pose.
Co-op Translator is an open-source translation tool from Microsoft that uses Azure AI to automate multilingual translation of text in project documents and images. With just one command, Co-op Translator can analyze project content...
AnchorCrafter is an intelligent video production system based on a diffusion model, used to automatically generate high-fidelity, anchor-style product promotion videos. It integrates human-object interaction (HOI) with gesture-guided human video generation...
MyTimeMachine (MyTM) is an advanced personalized facial age transformation technology that trains a personalized pre-trained global aging model using approximately 50 personal photos spanning 20 to 40 years. My...