The following joint laboratories are sorted according to their establishment time.

The following joint laboratories are sorted according to their establishment time.

1、 Tsinghua Hitachi Green Information Technology Joint Laboratory

2、 Tsinghua iFLYTEK joint research center

3、 Tsinghua Roma joint research center

4、 Tsinghua University car audio network auditory information processing joint research center

1、 Tsinghua Hitachi Green Information Technology Joint Laboratory

Established on October 21, 2001.

The main research contents are as follows

1. New paradigm for IPv6 based mobile Internet

2. QoS control for mobile multimedia services

3. Wireless sensor NW for always-be-connected

4. High-density WLAN for always-best-connected

5. ICBN for tri-network convergence

6. BS cooperation and DAS for LTE/LTE-A

7. Green communication and networks

8. Industrial radio technologies

Joint laboratory activities

2、 Tsinghua iFLYTEK joint research center

In 2006, the multimedia signal and Intelligent Information Processing Laboratory of the Department of electronic engineering formed a strategic alliance with iFLYTEK Information Technology Co., Ltd. of Anhui University of science and technology, and established Tsinghua iFLYTEK joint research center.

  1. About the joint research center

The purpose of the joint research center is to achieve a virtuous circle of industry university research through close cooperation between industry, University and Research Institute. Relying on the advanced scientific research level of Tsinghua University, it provides basic research support of artificial intelligence for iFLYTEK's products, and enhances the industry competitiveness and market leadership of iFLYTEK; at the same time, iFLYTEK creates a benign and interactive research environment for the joint research center to ensure that the research demand comes from the application and the research results serve the application; and jointly promote the artificial intelligence technology and market leadership The leap forward development of industrial application.

Since the establishment of the joint laboratory, the research direction has expanded from the initial intelligent voice and language technology to knowledge engineering, multimodal recognition, data mining and other directions; the research results have been widely used in medical care, intelligent customer service, voice chip and other application fields.

At present, the joint laboratory has a high-level research team composed of engineers and graduate students. The current director is Professor Wu Ji of the Department of electronic engineering.

2. Basic information and industry status of iFLYTEK

IFLYTEK Co., Ltd., established in 1999, represents the international advanced level in the core technologies of intelligent voice and artificial intelligence, such as speech synthesis, speech recognition, oral evaluation, language translation, voiceprint recognition, face recognition, natural language processing, etc. It was listed on Shenzhen Stock Exchange in 2008, stock code: 002230.

IFLYTEK is the only "industrialization base of national 863 Program achievements", "key software enterprises in national planning layout" and "national high-tech industrialization demonstration project" with voice technology as the industrialization direction in China. It has been determined as the leader unit of Chinese voice interaction technology standard working group by the former Ministry of information industry, and takes the lead in formulating Chinese voice technology standards. IFLYTEK was selected as one of the first batch of national new generation artificial intelligence open innovation platform, which explicitly relies on the company to build a new generation of intelligent voice National Open Innovation Platform of artificial intelligence. "National intelligent voice high tech industrialization base" and "voice and language information processing National Engineering Laboratory" have successively settled in the company, which is conducive to further gathering resources and improving the leading position of iFLYTEK industry.

With the advent of the mobile Internet era, iFLYTEK took the lead in releasing the world's first iFLYTEK open platform to provide intelligent voice interaction capability of mobile Internet, and continued to upgrade and optimize. Based on this platform, iFLYTEK has successively launched iFLYTEK input method, Migu Lingxi and other demonstration applications, and has worked with its partners to promote various voice applications to mobile phones, automobiles, household appliances, toys and other fields, leading and promoting the transformation of input and interaction mode in the era of mobile Internet.

In 2014, with the arrival of the era of artificial intelligence, iFLYTEK launched the "iFLYTEK Super Brain plan". The goal is to make the machine not only "listen and speak, but also" understand and think ", so as to realize a Chinese cognitive intelligent computing engine, which will lead to intelligent applications in the fields of home furnishing, education, customer service, medical treatment and so on. In 2016, the number of users of iFLYTEK's artificial intelligence open platform doubled with the number of entrepreneurial teams, driving more than one million people to carry out entrepreneurship and innovation activities. IFLYTEK's intelligent platform has been open for several times from 0.46 billion to 0.6 billion in 2018.

IFLYTEK official website:

Typical research results and industrialization of joint research center

1) Application of "intelligent medical assistant" medical examination robot in auxiliary medical field

In 2017, the "intelligent medical assistant" jointly developed by iFLYTEK and Tsinghua University became the first artificial intelligence robot in China and even in the world to pass the comprehensive written examination of national medical licensing examination with excellent scores exceeding 96 points. In essence, this is a natural language processing system that is particularly good at dealing with medical problems. The algorithm is to find the evidence needed to answer a specific question by calculating the statistical similarity between the words represented by mathematics. A detailed analysis of the test results shows that machines cannot compete with human beings in common sense and ethics. About three months after the results of the "intelligent medical assistant" were announced, the "intelligent medical assistant" officially entered the grass-roots hospitals to provide the general practitioners and residents with artificial intelligence assisted diagnosis and treatment services. At the same time, the "intelligent medical assistant" can quickly generate the electronic medical record and give the diagnosis of the disease according to the dialogue between the doctor and the patient. After seeing the diagnosis of the "intelligent medical assistant", the doctor can check the patient's medical history through the "intelligent medical assistant", and can also query similar cases, clinical guidelines and the drugs used for the symptomatic.

2) Application of natural language processing technology in precise medical text retrieval

Precision medicine is a new medical concept and medical model based on individualized medicine, with the rapid progress of genome sequencing technology and the cross application of bioinformatics and big data science. The diagnosis and treatment of patients are based on the biological characteristics of patients, phenotypes and social characteristics. Through the introduction of medical knowledge base, hierarchical rule optimization and PM document classifier based on deep neural network, the laboratory research group ranked first in two sub tasks of TREC 2018 precision medicine (PM) track in total six indicators, three of which ranked first and two ranked second, ranking first among all the teams.

3) Application of voice quality inspection and analysis technology in customer service center of mobile operators

Call center generates a large number of voice data every day. On the one hand, intelligent voice analysis technology can effectively improve the quality inspection efficiency of call center and improve user satisfaction; on the other hand, it can effectively analyze user feedback information to provide support for product design, marketing analysis and business management of enterprises. After the application of the results in the operator call center, the violation rate of the call center decreased by 87%; and the marketing activities were effectively analyzed, and the project results were highly recognized by the user units and are being promoted nationwide.

Speech analysis system

4) Application of semantic understanding technology in mobile SMS customer service

The e-channel intelligent service system integrates advanced natural language understanding technology to help users get rid of tedious business codes. Through natural language dialogue, the functions of business consultation, inquiry and processing can be realized. The online application of the system helps users to handle business through natural language, which leads to the increase of business customization; it increases the probability of mobile users using electronic channels by three times, effectively reducing the workload of manual customer service; the introduction of the system also increases the marketing mode of electronic channels, and operators can launch more flexible preferential activities to users through natural language understanding technology. At present, the system has been widely used in more than ten provincial operators in China and achieved good results.

Short message service system of electronic channel

3、 Tsinghua Roma joint research center

Rohm Semiconductor Co., Ltd. and Tsinghua University jointly established the Tsinghua Rohm joint research center on October 15, 2009. It has carried out extensive industry university cooperation in the fields of optoelectronics, biomedical engineering, power electronics, digital television and sensor networks. A series of achievements have been made.

Signing ceremony


  1. To become a platform for academic and cutting-edge application technology exchange between Tsinghua University and Roma company, effectively promote the "industry university research" alliance between the University and the company, create opportunities for Industry University exchange, facilitate the generation of cooperative projects, and promote the development of relevant academic research, new technology development and Application of its achievements.

2. Combine the advantages of University in academic research with the capabilities of Roma company in product development, and build a world-class research base in the field of semiconductor electronics relying on the superior hardware facilities of "Roma building".

3. Cultivate top-notch innovative talents in the field of semiconductor electronics.

Joint research center facilities

4、 Tsinghua University car audio network auditory information processing joint research center

Shenzhen car voice network technology Co., Ltd. is a high-tech enterprise with voice and audio processing technology and Internet of vehicles technology as the core. The joint research center for auditory information processing was established in December 2013. The joint research center will focus on the application requirements of human-computer interaction, information security, voice control, music retrieval, language learning, media computing, etc., and combine with the technical difficulties proposed by "car voice network" company to jointly tackle key problems, and innovate in methods, technologies and theories.

The main research contents are as follows

1. Engineering application optimization and joint technology research, improve processing speed and robustness, and put into practical application.

2. Study language recognition technology, speaker recognition technology and keyword recognition technology under noise and complex channel.

3. Research on the modeling method of humanized interactive dialogue system in vehicle environment.

4. Research on speech and audio processing methods of microphone array in vehicle environment..

5. Study the modeling method and theory of multilingual speech recognition.