AstrBot is a multi-platform chatbot and development framework that supports various large language models (such as OpenAI GPT, Google Gemini, Llama, etc.) and multiple messaging platforms (such as QQ, Telegram, WeChat, etc.). AstrBot...
Lumina-Image 2.0 is an open-source, high-efficiency unified image generation model with 2.6 billion parameters, based on a diffusion model and Transformer architecture. It excels in image generation quality, complex cue understanding, and resource efficiency.
BEN2 (Background Erase Network 2) is a deep learning model developed by Prama LLC, specifically designed for quickly removing backgrounds and extracting foregrounds from images and videos. It is based on an innovative confidence-guided matting (CGM) pipeline...
PartEdit is a fine-grained image editing method introduced by KAUST, based on a pre-trained diffusion model. PartEdit optimizes specific text markers (called "part markers"), allowing the diffusion model to accurately locate and edit various parts of objects within an image...
Agentic Object Detection is a novel object detection technology developed by Andrew Ng's team. It achieves object detection without labeled data through an intelligent agent system. Users only need to input text prompts, and the AI performs inference...
Agno is a lightweight framework for building intelligent agents. Agno supports multimodal (text, image, audio, video) and multi-agent collaboration, enabling rapid agent creation—up to 5000 times faster than LangGraph—and supports...
Ola is a multimodal language model developed collaboratively by Tsinghua University, Tencent Hunyuan Research Team, and the National University of Singapore's S-Lab. Through a progressive modality alignment strategy, it gradually expands the modalities supported by the language model, from images and text...
AlphaGeometry2 is an advanced artificial intelligence system developed by Google DeepMind, specifically designed to solve geometry problems in the International Mathematical Olympiad (IMO). It combines neural symbolic methods with Google's Gemini system...
EliGen is a novel entity-level controllable image generation framework jointly developed by Zhejiang University and Alibaba Group. By introducing a region attention mechanism, it can seamlessly integrate entity cues and spatial masks of arbitrary shapes without additional parameters...
Hibiki is an open-source decoder model from Kyutai Labs for simultaneous speech translation, capable of translating speech in one language into speech or text in another language in real time. Hibiki is based on a multi-stream language model architecture, processing the source language simultaneously...
HMA (Heterogeneous Masked Autoregression) is an open-source method developed by MIT, Meta, and the University of Illinois at Urbana-Champaign for modeling the dynamics of robot motion videos. HMA is based on heterogeneous pre-training and uses...
DeepRant (Chinese name: 鲸喷) is a multilingual translation tool designed specifically for gamers. DeepRant helps players quickly communicate via text on international servers, eliminating language barriers. Players select...
StochSync (Stochastic Diffusion Synchronization) is an innovative image generation technology specifically designed for generating high-quality images in complex spaces, such as 360° panoramas or 3D surface textures. It combines diffusion synchronization...
Sonic is an audio-driven portrait animation framework developed by Tencent and Zhejiang University. It generates realistic facial expressions and movements based on global audio perception. Sonic utilizes context-enhanced audio learning and a motion decoupling controller to extract audio...
FireRedASR is an open-source, industrial-grade Automatic Speech Recognition (ASR) model family from Xiaohongshu. It supports Mandarin, Chinese dialects, and English, achieving a new state-of-the-art (SOTA) level in Mandarin ASR benchmark tests, and also excelling in lyric recognition...
MVoT (Multimodal Visualization-of-Thought) is a novel multimodal reasoning paradigm developed by Microsoft Research, the Language Technologies Lab at the University of Cambridge, and the Institute of Automation, Chinese Academy of Sciences. It is based on generative image visualization...
DynVFX is an innovative video enhancement technology that seamlessly integrates dynamic content into real-world videos based on simple text commands. By combining a pre-trained text-to-video diffusion model and a visual language model (VLM), it achieves seamless integration of dynamic content into real-world videos...
WorldSense, developed by Xiaohongshu and Shanghai Jiao Tong University, is a benchmark test used to evaluate the comprehensive understanding capabilities of multimodal large language models (MLLMs) of visual, auditory, and text input in real-world scenarios. WorldSense...
Kiln AI is an open-source AI development tool that simplifies fine-tuning of large language models (LLMs), synthetic data generation, and dataset collaboration. Kiln AI offers an intuitive desktop application that supports Windows, macOS, and Linux...
Airweave is an open-source tool that can synchronize data from any application (including APIs, databases, websites, etc.) to graph and vector databases, enabling data retrieval based on intelligent agents or search mechanisms. Airweave...
MotionCanvas is an image-to-video (I2V) generation method developed by the Chinese University of Hong Kong, Adobe Research, and Monash University. It can transform still images into videos with rich dynamic effects. MotionCanvas is based on...
MedRAX (Medical Reasoning Agent for Chest X-ray) is a medical reasoning AI agent specifically designed for chest X-ray examinations. It integrates state-of-the-art chest X-ray analysis tools with a multimodal, large-scale language model to form a...
LangBot is an open-source instant chatbot platform that supports multiple platforms (such as QQ, WeChat, Lark, Discord, etc.) and various large language models (such as ChatGPT, DeepSeek, Gemini, etc.). LangBot features multi-modal...
LLMDet is an open-source lexical object detector developed by Alibaba Group's Tongyi Lab, the School of Computer Science and Engineering at Sun Yat-sen University, and Pengcheng Lab, among other institutions. It improves object detection based on co-training with large language models (LLMs)...