|Position Summary||We are developing advanced application technologies and intelligent information processing technologies to contribute to the variety of business domains of Sony Group Corporation, such as Agent Platform, Next-Gen Financial Front-end, Contents Creation Technology, Healthcare/Medical Service and Telepresence.
In cooperation with other departments, we are developing elemental technologies necessary for the development of dialog agent system, such as speech recognition, speech synthesis, user speech analysis, dialog control, and chatting. The developed elemental technologies can be provided as SaaS.
We also develop and provide SDKs to support the development of agent system with character visuals such as 2D animations and 3D graphics.
[Next-Gen Financial Front-end]
By introducing agent technology based on interaction in natural language into the front ends (customer touch points) of each financial GP of real estate, bank, non-life insurance, and life insurance, (part of) the business of a sales representative, a call center operator, a life planner, etc. is replaced by an agent, and a new front end is realized in which a contact point with a customer in financial service can be easily used in natural language.
[Contents Creation Technology]
For the entertainment business such as movies, animation, games, and music, we are developing automatic generation technology of dialogue and live comment with character, and production support technology of scenario, movie, and music. In addition to using these technologies to improve the efficiency of content creation workflows and to enhance the creativity of creators, we are also collaborating with related divisions and elemental technology development teams to develop automatic content creation for "virtual production" in the future.
In the field of healthcare and medical care, it is important to grasp the condition and QoL of patients in order to carry out appropriate diagnosis and treatment. At present, it is generally done by specimen examination (example: Blood tests) and inquiry and questionnaire (PRO) at medical institutions. While accurate information can be obtained, specimen examination can often be performed only by medical institutions, and only at that time can be observed. In addition, interviews and questionnaires are a burden on doctors and patients, and there is a lot of information based on patients' memories, so it can be said that the accuracy is somewhat lacking.
By utilizing Sony's interaction, sensing, and data analysis technologies to accurately understand the patient's condition and QoL in various situations during clinical trials and daily life, and by combining them with current methods, it is possible to make better use of them in diagnosis and treatment. As the spread of telemedicine accelerates in the future, it will become more and more important to accurately grasp patient conditions and QoL. While developing technologies for this purpose, we aim to expand it to telemedicine and remote monitoring in the medium term.
A telepresence system that connects people to people, people to space, and allows "natural communication as if you were in the same space," regardless of distance limitations.
The main work is research and development of advanced algorithms using software. It includes tasks such as artificial intelligence, deep learning, natural language processing, UI design and mobile, cloud service implementation, 3D graphics, technical evaluation and analysis, and architecture design.
As an R&D engineer, you will work on a specific project to develop integration technologies to enhance customer value, aiming to contribute to the variety of business of Sony Group Corporation.
We are looking for R&D engineers who have keen interest and ability to apply deep learning technologies to real world problems.
|Responsibilities||Research and development of advanced application technologies and information processing technologies in our target fields. (Computer Vision, Natural Language Processing, Sound, etc.)
- Open-domain dialogue generation.
- Vision-based commentary generation.
- Language-based procedural CG generation.
- Non-verbal context understanding.
- CG Avatar / Motion generation.
- Multi-turn dialogue understanding.
- Real-world data analysis.
- Respiratory / Heart sound analysis.
- Behavioral change interventions/systems/theory in healthcare.
|Required qualifications||- Master's degree in Computer Science.
- 2 years of research experience in Machine Learning fields. (Vision, Language, and Graphics.)
- Collaboration, teamwork and knowledge sharing within the group.
- Interest and ability to bring creative, non-standard, and innovative solutions to the table.
- Working proficiency and communication skills in verbal and written English.
|Preferred qualifications||- PhD, or higher qualification in Computer Science.
- Research experience in Algorithms, Application, Architecture, AI, Database, Data Mining, Machine Learning, Cloud Service, User Interface.
- Experience applying Machine Learning to real world problems.
- Willingness in learning Japanese.
|Product, Service||Entertainment Contents, PlayStationÂ®, Robots, Financial Services, Healthcare Services, Remote Communication, and Consumer Electronics.|
|Development Environment||- Platform: Windows, Linux, AWS, GCP, etc.
- Computing: Private GPU Cluster for Deep Learning.
- Communication: GitHub, Slack, Teams, etc.
|Coding Test||Not Required|