Jane Shen specializes in AI, deep learning, face, person and image recognition, autonomous driving with more than 20 years of experience. She has known-how and domain knowledge across different applications from camera products, surveillance, smart city to autonomous driving. She is currently a chief scientist and managing director at Pensees Singapore. She is working on cutting edge technology development by embracing AI and deep learning advancement with strength in computer vision.
Before joining Pensees, Jane was an assistant director at Panasonic R&D Center Singapore and general manager at Panasonic Singapore Lab, from 2012 to 2019. During this time, she led a team in autonomous development with AI deep learning technology. Under her leadership, the team developed several face recognition products that won in several competitions, including World No.1 position in NIST IJB-A and MS-Cele-1M Face Challenge in 2017 and 2018, and ICCV 2019 Face recognition challenge, achieving world top position for Person reID, car reID, fighting and AI City Challenge.
She has wide range of experience in audio/video compression and coding, digital rights management, camera and sensor processing, computer vision, which are applied in many domains, such as AV products, securities and surveillance, DTV/DVD, automotive, sensor device, content authoring, delivering and receding, etc. In addition, she has filed more than 300 patents. She is proficient in system and architecture design, especially in system solution.
A big scale AIoT system was developed and commercialized for a city in China. It is a cloud based CCTV video analytics system with edge box and video analytics engines, where the face recognition and person re-identification AI engines was deployed with other technologies, including face detection, tracking and quality assessment.
- Infocomm - Test & Verification/Clinical Trials
- Electronics - Test & Verification/Clinical Trials, Product Design & Development
Most CCTV cameras are passively used today, and incident discovery is still highly demanded for many restricted places. However, manpower resources are not sufficient to cater for these tasks.
A project was completed in Singapore, in which a set of the designed AI technologies are integrated in a system to automatically monitor the scene by analysing the video streams from cameras. Abnormal behaviors or situations, including fighting/aggressiveness will be identified and alerts will be sent to persons in charge via phone and PC, allowing them to take action immediately. This can not only save the manpower but also to quickly and automatically identify incidents which might trigger dangers.
- Infocomm - Certification/Regulatory Compliance
AI based CCTV system is highly demanded for premises owners or security agents to improve the productivity upon limited manpower resources.
AI based Visual Analytics (VA system (i.e., 9 VAs with VA server) has been designed and developed in several clusters of buildings, to provide intelligent VA functions and automatically sending alert to business owners.
- Infocomm - Product Design & Development