FAQIVS: Face Query-based Interactive Video Synopsis*
Synoptic video is an intelligent and short video that provides vital content according to user requirements. It would reduce human resources and time requirements and alleviate the video browsing burden. “Face” is the primary authenticated attribute to confirm the victim. The challenges include low-...
Saved in:
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Taylor & Francis Group
2025-04-01
|
| Series: | Automatika |
| Subjects: | |
| Online Access: | https://www.tandfonline.com/doi/10.1080/00051144.2025.2459987 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Synoptic video is an intelligent and short video that provides vital content according to user requirements. It would reduce human resources and time requirements and alleviate the video browsing burden. “Face” is the primary authenticated attribute to confirm the victim. The challenges include low-resolution, tiny faces, and partial occlusion. The significant contribution is to create a novel face-query-based interactive video synopsis (FaQIVS) and the background and timestamp in different zones of field-of-view. The proposed synoptic video algorithm considers the challenges such as the presence of the same (Query) person at different zones the repeated presence of the same person at the same field-of-view zone at different times, the field-of-view at different times, and the repeated presence of the same person at same field-of-view zone at a same/different day/time. The proposed FaQIVS synopsis provided an F-score of 95% on the Chokepoint dataset and KCET dataset over state-of-the-art methods. |
|---|---|
| ISSN: | 0005-1144 1848-3380 |