The Realtime API, introduced by OpenAI, is a low-latency, multimodal conversational API that supports text and audio as input and output. The Realtime API allows developers to build near real-time interactive experiences, such as voice-to-voice...
Laminar is an open-source observability and analytics platform designed for large language model (LLM) applications. Laminar provides a complete suite of tools for tracking, evaluating, annotating, and analyzing LLM data, enabling developers to gain deeper insights into...
MIP-Adapter is a personalized image generation technology launched and open-sourced by Alibaba Group. Based on the IP-Adapter model, it further expands its capabilities to support the simultaneous processing of multiple reference images, generating more accurate and high-quality images...
CapsWriter-Offline is a high-efficiency offline voice input and text transcription tool for PCs, allowing users to achieve real-time speech-to-text conversion through simple keystrokes. The software is suitable for quickly recording and transcribing large amounts of voice information...
HelloBench is an open-source benchmark for evaluating the long text generation capabilities of large language models (LLMs). HelloBench includes five subtasks based on Bloom's Taxonomy: open-ended question answering, summarizing, chat, text...
VirtualWife is an innovative virtual digital human project focused on developing AI-powered virtual characters. These characters are designed for interaction on live streaming platforms like Bilibili, providing entertainment and information. By integrating advanced AI technologies, such as...
MMSearch is a benchmark for evaluating the capabilities of large multimodal models (LMMs) as AI search engines. It includes an MMSearch-Engine framework and the MMSearch test suite, which contains 300 questions covering 14...
Anime.gf is an open-source native application that allows users to create and interact with their own virtual avatars. The tool provides a user-friendly interface, enabling users to communicate with a variety of virtual avatars with unique personalities and language styles.
ChopperBot is an intelligent live video editing and publishing robot that automates the management and distribution of live content. It supports multiple popular live streaming platforms, such as Douyu, Huya, Bilibili, Douyin, and Twitch,...
FineZip is a lossless text compression system based on Large Language Models (LLMs). It combines online memory and dynamic context size techniques to improve the speed and efficiency of text compression. Online memory refers to the process of modifying the model before compression...
Depth Pro is an advanced monocular depth estimation model developed by Apple that can quickly generate high-resolution 3D depth maps from a single 2D image. The model is not only fast, taking only 0.3 seconds, but also provides metric-level depth information...
VideoLingo is a one-click, fully automatic video translation tool that can segment, translate, align, and add audio to videos, ultimately generating Netflix-quality subtitles and audio. VideoLingo is based on Natural Language Processing (NLP)...
CatVTON is an advanced virtual try-on technology jointly developed by Sun Yat-sen University and Pixocial. Based on a lightweight architecture and efficient training strategies, it achieves high-quality virtual try-on results. A key feature of CatVTON is that it requires minimal...
ScribbleDiff is an advanced text-to-image generation technique that uses visual cues from simple user doodles to guide the image generation process. It analyzes the doodles to ensure that the orientation of objects in the generated image matches the orientation of the user's doodles...
Boow-VTON is an advanced virtual try-on technology launched by Alibaba. It enables high-quality virtual try-on effects in outdoor settings without relying on precise masking or retouching. Through effective data augmentation methods, based on...
LiveKit Agents is a powerful framework for creating multimodal AI agents that can interact with users in real time via voice, video, and data. The framework supports Python programming, simplifying the development process and allowing developers to easily integrate with it...
Matryoshka Diffusion Models (MDM) are an innovative diffusion model introduced by Apple, primarily used for generating high-resolution images and videos. MDM utilizes a multi-resolution diffusion process, simultaneously performing diffusion at different scales...
IFAdapter is a novel text-to-image generation model jointly developed by Tencent and the National University of Singapore. It improves the accuracy of location and feature generation when generating images containing multiple instances. Traditional models struggle with multi-instance images...
TinyVLA is a vision-language-action (VLA) model for robot manipulation, developed by a team from East China Normal University and Shanghai University. It addresses the shortcomings of existing VLA models, such as slow inference speed and the need for large amounts of pre-trained data...
Inverse Painting is an AI technology developed by researchers at the University of Washington that reverses the painting process. It learns painting techniques and sequences by analyzing videos of artists painting, and then generates a series of painting instructions...
Playground v3 (PGv3) is the latest text-to-image model from Playground Research. Based on deep fusion Large Language Model (LLM) technology, it surpasses the capabilities of human designers in graphic design tasks...
Reverb ASR is an open-source automatic speech recognition and speaker separation model from Rev. Inc., trained on 200,000 hours of manually transcribed English data. The model excels in long speech recognition and is suitable for processing tasks such as podcasts and financial statements...
Open NotebookLM is an open-source AI tool that uses the latest open-source AI models, such as Llama 3.1 405B, MeloTTS, and Bark, to convert PDF documents into podcast-style audio content. The tool is suitable for converting written information...
StoryDiffusion is an advanced AI image and video generation framework for generating consistent image and video sequences from text descriptions. It enhances image consistency based on a Consistent Self-Attention mechanism...