xAR is a novel autoregressive visual generation framework jointly proposed by ByteDance and Johns Hopkins University. The framework utilizes "Next-X Prediction" and "Noisy Context Learning"...
Avat3r is a high-fidelity, large-scale, animable Gaussian reconstruction model for 3D head images, developed by the Technical University of Munich and Meta Reality Labs. It requires only a few input images to generate high-quality, animable 3D head images...
WhisperChain is an open-source speech recognition tool that improves work efficiency based on voice input. WhisperChain uses Whisper.cpp to implement real-time speech recognition, converting speech to text, and then uses LangChain to process the text...
Spark Medical Large Model X1 is a deep reasoning large-scale model released by iFlytek, specifically designed for the medical field and possessing powerful capabilities for handling complex problems. Through deep reasoning algorithms, the model can progressively explain the evidence-based process, significantly reducing...
Story-Flicks is an AI-based project that supports one-click generation of high-definition short story videos. After the user inputs a story theme, the system uses AI technology to generate a short video containing images, text, audio, and subtitles. Story-Flicks...
CSM (Conversational Speech Model) is a new voice dialogue model developed by the Sesame team, enhancing the naturalness and emotional interaction capabilities of voice assistants. CSM is based on a multimodal learning framework, combining text and speech data...
AgiBot Digital World is a high-fidelity robot simulation framework developed by AIZOO Robotics, providing efficient support for robot operation skills research and application. AgiBot Digital World integrates a massive amount of realistic 3D assets, multiple...
ARTalk is a novel voice-driven 3D head animation generation framework developed by the University of Tokyo and RIKEN. Based on an autoregressive model, it achieves real-time, highly synchronized lip movements and natural facial expressions and head poses. A...
LuminaBrush is an interactive tool for painting lighting effects on images. Based on the Flux texturing project, LuminaBrush uses a two-stage approach: the first stage transforms the image into a "uniformly lit" appearance, and the second stage...
HumanOmni is a large, multimodal model focused on human-centric scenarios, fusing visual and auditory modalities. By processing video, audio, or a combination of both inputs, it comprehensively understands human behavior, emotions, and interactions. The model is based on...
Spark-TTS is an open-source, high-efficiency text-to-speech (TTS) tool based on Large Language Models (LLMs) from the SparkAudio team. It reconstructs audio directly from LLM-predicted encodings without requiring additional generative models, achieving zero-sample...
ViDoRAG is a visual document retrieval enhancement and generation framework developed by Alibaba Tongyi Lab in collaboration with the University of Science and Technology of China and Shanghai Jiao Tong University. Based on multi-agent collaboration and dynamic iterative reasoning, it addresses the challenges of traditional methods in handling complex...
Shandu is an open-source AI research automation tool that combines LangChain and LangGraph technologies to automate multi-level information mining and analysis, generating structured research reports. Shandu's core functionality...
LCVD (Lighting Controllable Video Diffusion Model) is a high-fidelity, lighting-controlled portrait animation generation framework developed by Sichuan University. LCVD is based on separating the intrinsic features of a portrait (such as identity and appearance) from...
SepLLM is an efficient framework for accelerating Large Language Models (LLMs), jointly proposed by the University of Hong Kong, Huawei Noah's Ark Lab, and other institutions. It significantly improves the inference speed of the model by compressing paragraph information and eliminating redundant tags...
CogView4 is an open-source text-to-image model developed by Zhipu, featuring 6 billion parameters and supporting native Chinese input and Chinese text generation. The model ranked first in the DPG-Bench benchmark test, achieving top scores among open-source text-to-image models...
PRefLexOR (Preference-based Recursive Language Modeling for Exploratory Optimization of Reasoning) is a novel self-learning AI framework proposed by an MIT team, combining preference optimization...
Probly is an AI-powered spreadsheet tool that combines spreadsheet functionality with Python's data analysis capabilities. Based on WebAssembly, Probly runs Python code in a browser and supports interactive spreadsheets, data processing, and more...
MindLLM is an AI model jointly developed by Yale University, Dartmouth College, and Cambridge University. It decodes functional magnetic resonance imaging (fMRI) signals from the brain into natural language text. MindLLM is based on a subject-independent...
MiniMind is an open-source, ultra-small language model project that helps individual developers train their own language models from scratch at extremely low cost. MiniMind is designed for lightweight use; the smallest version requires only 25.8M parameters and is only a GPT-...
Fractal Generative Models are a novel image generation method developed by the MIT Computer Science and Artificial Intelligence Laboratory and Google DeepMind. Fractal Generative Models...
Image-01 is an advanced text-to-image generation model from MiniMax, boasting exceptional image generation capabilities. It accurately converts user-input text descriptions into high-quality images, supporting various aspect ratios and high-resolution outputs...
SuperGPQA is a knowledge reasoning benchmark test suite jointly launched by ByteDance's Doubao Big Model Team and M-A-P. It comprehensively covers 285 graduate-level disciplines and contains 26,529 professional questions. It addresses the limitations of traditional benchmark testing in terms of subject coverage...
SpeciesNet is an open-source artificial intelligence model from Google that identifies animal species by analyzing photos taken with camera traps. Trained on over 65 million images, it can recognize over 2,000 tags, including animals...