A Ulam's game for video based action recognition
In this dissertation, we propose a conditional early exiting framework with Ulam’s Game for action recognition. Since the action recognition system has extremely high requirements on dynamic performance, our system pays more attention to improving the detection efficiency of the system, hoping to ob...
| Main Author: | |
|---|---|
| Other Authors: | |
| Format: | Thesis-Master by Coursework |
| Language: | English |
| Published: |
Nanyang Technological University
2022
|
| Subjects: | |
| Online Access: | https://hdl.handle.net/10356/161732 |
| _version_ | 1824453235841171456 |
|---|---|
| author | Zheng, Haofeng |
| author2 | Tay Wee Peng |
| author_facet | Tay Wee Peng Zheng, Haofeng |
| author_sort | Zheng, Haofeng |
| collection | NTU |
| description | In this dissertation, we propose a conditional early exiting framework with Ulam’s Game for action recognition. Since the action recognition system has extremely high requirements on dynamic performance, our system pays more attention to improving the detection efficiency of the system, hoping to obtain the detection results in a shorter time. In our system, we use a modified ResNet-50 as backbone network to do feature extraction and use a Pooling module to accumulate feature. Then, we have a neural network Gate module to determine whether the feature have accumulated enough to begin Ulam’s Game. A classifier is used to get candidate results, which are used to run Ulam’s Game and get the final prediction. The model shows good detection accuracy and dynamic performance in multiple data sets (Mini-Kinetics, ActivityNet). |
| first_indexed | 2024-10-01T07:57:31Z |
| format | Thesis-Master by Coursework |
| id | ntu-10356/161732 |
| institution | Nanyang Technological University |
| language | English |
| last_indexed | 2024-10-01T07:57:31Z |
| publishDate | 2022 |
| publisher | Nanyang Technological University |
| record_format | dspace |
| spelling | ntu-10356/1617322022-09-19T06:02:36Z A Ulam's game for video based action recognition Zheng, Haofeng Tay Wee Peng School of Electrical and Electronic Engineering wptay@ntu.edu.sg Engineering::Electrical and electronic engineering In this dissertation, we propose a conditional early exiting framework with Ulam’s Game for action recognition. Since the action recognition system has extremely high requirements on dynamic performance, our system pays more attention to improving the detection efficiency of the system, hoping to obtain the detection results in a shorter time. In our system, we use a modified ResNet-50 as backbone network to do feature extraction and use a Pooling module to accumulate feature. Then, we have a neural network Gate module to determine whether the feature have accumulated enough to begin Ulam’s Game. A classifier is used to get candidate results, which are used to run Ulam’s Game and get the final prediction. The model shows good detection accuracy and dynamic performance in multiple data sets (Mini-Kinetics, ActivityNet). Master of Science (Computer Control and Automation) 2022-09-19T06:02:36Z 2022-09-19T06:02:36Z 2022 Thesis-Master by Coursework Zheng, H. (2022). A Ulam's game for video based action recognition. Master's thesis, Nanyang Technological University, Singapore. https://hdl.handle.net/10356/161732 https://hdl.handle.net/10356/161732 en ISM-DISS-02979 application/pdf Nanyang Technological University |
| spellingShingle | Engineering::Electrical and electronic engineering Zheng, Haofeng A Ulam's game for video based action recognition |
| title | A Ulam's game for video based action recognition |
| title_full | A Ulam's game for video based action recognition |
| title_fullStr | A Ulam's game for video based action recognition |
| title_full_unstemmed | A Ulam's game for video based action recognition |
| title_short | A Ulam's game for video based action recognition |
| title_sort | ulam s game for video based action recognition |
| topic | Engineering::Electrical and electronic engineering |
| url | https://hdl.handle.net/10356/161732 |
| work_keys_str_mv | AT zhenghaofeng aulamsgameforvideobasedactionrecognition AT zhenghaofeng ulamsgameforvideobasedactionrecognition |