Multimodal Biometric Techniques for Forensics Tools
| Reference number | |
| Coordinator | KUNGLIGA TEKNISKA HÖGSKOLAN |
| Funding from Vinnova | SEK 300 000 |
| Project duration | September 2018 - September 2019 |
| Status | Completed |
| Venture | Personal mobility between societal sectors |
| Call | 2017-04722-en |
Important results from the project
The objective of this project is to renew the cooperation between KTH and NFC. NFC will gain from cutting edge research on biometrics and KTH professors gain from the exposure to real-world problems. The purpose is to study how to develop multimodal person-identification tools based on an integration of face, iris and fingerprint information into the body with performance markedly better than existing person identification tools in Swedish forensics. The goal will be that the tools to be developed would allow for improved identification of suspects in video surveillance videos.
Expected long term effects
1. A clear picture of existing and future challenges of person identification in video surveillance. 2. A comprehensive picture of what has been done previously around the topics and state of the art person identification techniques in video surveillance 3. Critical reviews on related information sources. 4. A good scientific foundation will be established for multimodal person identification. 5. A technical roadmap of developing multimodal person identification techniques into forensic tools. 6. A joint project on forensic face recognition has been initialised and funded by NFC
Approach and implementation
Professor Haibo Li worked on the NFC project for three months during the period: September 2018 to September 2019. Professor Li worked with Dr Peter Bergström and his research team in the NFC. A joint project team with representatives from NFC and KTH was established. On the NFC´s side, L Klasen, E Leitet, and P. Bergström have been active in the project. Three workshops took place in the NFC during the project period on multimodal biometrics, data-driven forensics, deep learning networks and human face identification, which attracted forensic scientists and NFC security practitioners.