MetaMorph is a multimodal large model (MLLM) that achieves visual understanding and generation through instruction tuning. It proposes a technique called Visual-Predictive Instruction Tuning (VPiT)...
AgentScope is an open-source multi-agent development platform from Alibaba Group, helping developers easily build and deploy multi-agent applications. AgentScope offers high ease of use, high robustness, and distributed support, and includes various built-in model APIs...
LeviTor is an image-to-video synthesis technology developed by institutions such as Nanjing University, Ant Group, and Zhejiang University. It combines depth information and K-means clustering points to control the trajectory of 3D objects in a video, eliminating the need for explicit 3D trajectory tracking. LeviTor...
3D-Speaker is a multimodal open-source project launched by the Speech Team of Alibaba Tongyi Lab. It achieves high-precision speaker recognition and language identification by combining acoustic, semantic, and visual information. 3D-Speaker provides industrial-grade...
ERA-42 is a large-scale, end-to-end native robot model developed by Beijing Xingdong Era. Combined with its self-developed five-finger dexterous hand, Xingdong XHAND1, it can perform over 100 complex and dexterous maneuver tasks. ERA-42 requires no pre-programming and possesses the ability to quickly learn new...
Baichuan4-Finance is a full-chain augmented model for the financial domain launched by Baichuan Intelligence, including Baichuan4-Finance-Base and Baichuan4-Finance. Based on Baichuan4-Turbo, it is trained using domain-specific self-constraints...
VSI-Bench (Visual-Spatial Intelligence Benchmark) is a visual-spatial intelligence benchmark set developed by Fei-Fei Li, Sai-Ning Xie, and their research team. The researchers constructed it to evaluate large-scale multimodal language models...
Bamba-9B is a decoding-only language model based on the Mamba2 architecture, jointly developed by IBM, Princeton University, Carnegie Mellon University, and the University of Illinois at Urbana-Champaign. The model is trained on completely open datasets and can improve...
Univer is an open-source, full-stack framework that supports the creation and editing of spreadsheets, documents, and presentations, providing users with a unified and powerful office solution. Univer runs in browsers and Node.js environments and is easily integrated into various...
Kheish is an open-source, multi-agent orchestration platform based on Large Language Models (LLMs). It coordinates the various steps of complex tasks, such as proposal generation, review, validation, and..., using multiple specialized roles (agents) and flexible workflows.
HelloMeme is a framework based on the latest Diffusion generation technology, enabling the transfer of facial expressions and poses. HelloMeme integrates a Spatial Knitting Attentions mechanism, based on Stable Diffusion...
InvSR is an innovative image super-resolution model that recovers high-resolution images based on the inverse process of a diffusion model. It improves super-resolution performance by leveraging rich image priors from large pre-trained diffusion models. The core of InvSR lies in depth...
CLEAR is a novel linear attention mechanism introduced by the National University of Singapore that improves the efficiency of pre-trained Diffusion Transformers (DiTs) in generating high-resolution images. Based on limiting the attention of each query to a local window, CLEAR...
DisPose is a control technology developed jointly by research teams from Peking University, University of Science and Technology of China, Tsinghua University, and Hong Kong University of Science and Technology to improve the quality of character animation. It is based on extracting effective data from skeletal pose and reference images...
OCTAVE (Omni-Capable Text and Voice Engine) is a next-generation speech and language model from Hume AI, combining the capabilities of the EVI 2 model with systems from OpenAI, Elevenlab, and Google DeepMind. OCTAVE...
Granite 3.1 is IBM's next-generation language model, boasting powerful performance and extended context processing capabilities. The Granite 3.1 model family includes four different sizes and two architectures: a dense 2B model and an 8B parametric model...
Sketch2Sound is an AI audio generation technology developed by Adobe Research and Northwestern University. It generates high-quality sound effects based on sound imitation and text prompts. Sketch2Sound extracts loudness, brightness, and pitch from sound imitation...
OmniAudio-2.6B is an audio language model from Nexa AI, designed for edge deployment, enabling fast and efficient audio-to-text processing. OmniAudio-2.6B is a multimodal model with 260 million parameters, fused with Gemma-2-2...
DreamOmni is a unified image generation and editing model jointly developed by the Chinese University of Hong Kong, ByteDance, and the Hong Kong University of Science and Technology. The model integrates text-to-image (T2I) generation with various editing tasks, including imperative editing, restoration, etc.
QVQ is an open-source multimodal reasoning model built by Alibaba based on Qwen2-VL-72B. It combines visual understanding and complex problem-solving capabilities to enhance the cognitive abilities of artificial intelligence. QVQ demonstrates enhanced capabilities in visual reasoning tasks, especially...
Open Notebook is an open-source, privacy-focused alternative to Google NotebookLM that helps users manage research workflows, generate AI-assisted notes, and interact with content. Open Notebook supports multiple notebooks and multiple models...
PC Agent is an advanced AI system jointly developed by Shanghai Jiao Tong University and the Generative AI Research Lab (GAIR). Based on simulating human cognitive processes, the system performs tasks such as organizing research materials, drafting reports, and creating presentations...
Midscene.js is an AI-based automation SDK that simplifies UI automation testing by using Large Language Models (LLMs). Users describe interaction steps or expected data formats in natural language, and Midscene.js executes...
ASAL (Automated Search for Artificial Life) is a system that uses a basic model to automatically search for artificial life (ALife). ASAL was jointly developed by Sakana AI, MIT, OpenAI, and other institutions, and is primarily based on three...