Hubei artificial intelligence (in Wuhan, artificial intelligence can output through "feeling&qu

1. Artificial Intelligence Image Robot

Many software have implemented the "face search" function, which uses facial recognition to find photos of the same person in the album. However, have you tried "text search" and sent text instructions to the computer to search for specific images? Can computers automatically generate visuals and scenes based on the text of the script by generating videos through text? Can computers create original songs based on the Tang and Song poetry of "using literature to create sound", allowing you to have a "classic chant and transmission" of thousands of people and songs?

2. Artificial intelligence image generation

Yes! In Wuhan, the multimodal big model can help AI, like the human brain, mobilize all "senses" to "feel" and output. Recently, the Wuhan Institute of Artificial Intelligence announced that it is working with the Institute of Automation of the Chinese Academy of Sciences and Huawei to build the "Zidong Taichu" 2.0 full mode big model, which is an upgraded version of 2.0 based on the world's first 100 billion parameter three mode big model, "Zidong Taichu" 1.0. In the future, text Unified representation and learning of different modalities such as images, audio, video, 3D, etc.

3. High definition artificial intelligence image materials

This provides Wuhan with unlimited imagination for the development of artificial intelligence applications. Last year, Wuhan received national support to create the "National Artificial Intelligence Innovation and Application Pilot Zone" at the beginning of this year, In the Implementation Plan for Building a National Artificial Intelligence Innovation and Application Pilot Zone in Wuhan (2023-2025) (Draft for Soliciting Opinions), it is explicitly proposed that Wuhan should "build a highland for innovation in large model applications" and "form a globally leading cross modal understanding and generation capability".

4. Artificial intelligence image creativity

What exactly is a 'big model'? The current hot ChatGPT is based on OpenAI's GPT-3.5 series of large-scale language models. AI applications based on big models have changed the way users use data. In the future, any data generated by enterprises or individuals can be used to train products of big models, bringing relevant value to enterprises.

5. Artificial intelligence image materials

According to incomplete statistics, up to now, there have been more than 40 companies and institutions in China that have released large model products or plans. Among these companies, there are many leading enterprises such as Baidu, Alibaba, ByteDance, Huawei, Xiaomi, etc., and most of them have a seat in Wuhan in Beijing, Shenzhen, Shanghai, Hangzhou, etc.

6. Artificial intelligence image high-definition

The winner of the "Outstanding AI Leader Award", the highest prize of the 2022 World AI Conference, and the world's first three mode big model "Zidong Taichu" were jointly developed by the Institute of Automation of the Chinese Academy of Sciences and Huawei in Wuhan. This big model applied the computing power of the Wuhan AI Computing Center and bred the Wuhan AI Research Institute in Wuhan.

7. Artificial Intelligence Picture Cartoon

Wang Jinqiao, Executive Deputy Director of the "Zidong Taichu" Large Model Research Center of the Institute of Automation of the Chinese Academy of Sciences and President of Wuhan Artificial Intelligence Research Institute, introduced that Wuhan proposes to build an industrial innovation ecosystem based on the support of large models for industrial quality inspection, film and television creation, smart healthcare, internet recommendation, intelligent driving and other scenario applications, which is exactly what the "Zidong Taichu" large model excels at.

One of the methods for artificial intelligence images is to use object detection

For example, in the field of industrial testing, manual detection used to be subjective, inefficient, and unable to monitor in real time. Using AI detection algorithms can enhance the objectivity of detection and improve detection efficiency. Real time detection promotes real-time correction in legal case analysis. Through the "Zidong Taichu" multimodal large model, a traffic accident case analysis, including event extraction and responsibility determination, only takes about 0.5 seconds, It can improve the efficiency of legal case analysis by about 100 times and has been applied to relevant legal enterprises in Wuhan.

One of the methods for artificial intelligence film reading is to use object detection

For another example, in the field of intelligent transportation, the use of multimodal perception data and the signals controlled at each intersection can optimize and make decisions on the three mode data of the road network, increase the road utilization rate by 30%, and help Wuhan's intelligent transportation to be upgraded again, The multimodal artificial intelligence industry consortium created by Wuzhi Academy and the Institute of Automation of the Chinese Academy of Sciences has gradually attracted nearly 70 member units from various industries, universities, and research institutes to jointly explore the industrialization path of universal artificial intelligence.

The process of artificial intelligence film reading is reflected in

Editor: Nie Yang [Source: Wuhan Municipal People's Government Website Wuhan News] Statement: The copyright of this article belongs to the original author. If there is any source error or infringement of your legitimate rights and interests, you can contact us through email, and we will handle it promptly. Email address: jpbl@jp.jiupainews.com .