Job Overview
- 採用企業タイプ 人材紹介・派遣・委託
- 業種 エンジニアリング, インフォメーションテクノロジー, インターネット
- Salary ¥5,000,000 / 年
- 英語レベル ビジネスレベル
- 日本語レベル ビジネスレベル
- 国内の応募者に限定? 国内在住応募者のみ
- ビザスポンサーシップ Yes
職務要件
- Someone with JLPT Level 2 Japanese language ability or higher
(MUST be able to adequately communicate in the workplace and understand work responsibilities in Japanese) - This position is needed to be the one who lives in Japan at the moment, not accepting applications from overseas
Required Skills and Experience
- Advanced programming skills in Python, C, C#, and open-source deep learning libraries (Pytorch/Tensorflow/Keras)
- Experience in developing applications/programs using deep learning libraries for image and speech analyses
- Experience in the implementation of deep learning methods (e.g. LSTM, RNN) for structured/temporal data analyses and imitation learning methods (e.g. IRL, GAIL)
- Programming skills in Unity/Unreal/Gazebo, and Robot Operating System (ROS)
- Programming experience in both Linux and Windows OS
- Experience in developing XR/VR applications, 3D reconstruction methods using computer vision (e.g. COLMAP, SfM, MVS pipeline) are added advantages
- Basic English/Chinese communication abilities are added advantages
職務内容
- Development of systems/programs for communication agents/robots in the physical and simulation environments based on the SENSE-THINK-ACT paradigm
- Experimental data collection and analyses
Main duties
- Development of programs/systems for real-time sensing/ recognition of human activities (e.g. facial recognition and body key points detection) and objects detection and recognition using multi-modal data (e.g. image and audio)
- Development of programs/systems for robot/agent behavioral control/learning and action generation in the physical and simulation environments
- Development of program/systems for 3D reconstruction of communication robot’s surrounding environment from 2D images
- Assist in experiment data collection and analyses
Personal qualities
- Able to work as a team, and to understand and use programs written by team members
Preferred start date
- As soon as possible