# Visual Recognition

UI-TARS-desktop is a desktop client application developed by ByteDance, built on the UI-TARS visual language model, which enables users to interact with computers using natural language to accomplish various tasks. This product leverages advanced visual language model technology to understand users' natural language commands and execute precise mouse and keyboard operations through screen capture and visual recognition. It supports cross-platform use (Windows and macOS) and provides real-time feedback and status updates, significantly enhancing user efficiency and interaction experience. The product is currently open source on GitHub, allowing users to download and use it for free.

Ideal Classmate is an AI application meticulously crafted by Ideal Automotive, leveraging its self-developed large model to provide users with a constantly available smart assistant. It possesses knowledge Q&A capabilities, answering questions across various fields such as automobiles, travel, finance, and technology, and is proficient in English translation and text generation to support users in learning and daily life. Additionally, it features visual perception abilities, allowing users and their families to explore the world and recognize various things encountered during outings. The product interface is designed to be simple and elegant, with precise and rapid voice input, and natural, smooth output that closely resembles human speech, making it an intelligent assistant that integrates knowledge Q&A, visual recognition, and voice interaction.

Lloyd is an application that integrates artificial intelligence visual recognition technology, capable of unlocking knowledge, insights, and information through video capture to enhance the video communication experience for users anytime and anywhere. Lloyd's main advantage lies in its real-time visual AI capability, allowing users to learn and discover new things instantaneously. Additionally, Lloyd offers fast video communication services, combining the convenience of text and voice messaging with the intuitiveness of video chatting.

Llama-3.2-90B-Vision is a multimodal large language model (LLM) released by Meta, focusing on visual recognition, image reasoning, image description, and answering general questions about images. The model surpasses many existing open-source and closed multimodal models in common industry benchmarks.

LLaVA-OneVision is a large multimodal model (LMM) collaboratively developed by ByteDance and several universities. It pushes the performance boundaries of open large multimodal models across single images, multiple images, and video scenarios. The model's design facilitates powerful transfer learning across different modalities/scenarios, showcasing new integrated capabilities, particularly in video understanding and cross-scenario abilities, demonstrated through task conversion from images to videos.

Unitor.ai is a personal voice and visual assistant that offers natural, warm voice conversations suitable for all ages and interests. It becomes smarter with every interaction, helping users organize their lives, provide emotional support and advice, as well as hands-free assistance while driving or working.

NavAIGuide is an extensible multimodal intelligent framework that automates tasks within mobile and desktop ecosystems by accessing applications. It features visual task detection, advanced code selectors, action-oriented execution, and robust error handling. NavAIGuide aims to provide users with efficient automation solutions.