End-to-end Active Object Tracking with Siamese Networks via Reinforcement Learning
碩士 === 國立交通大學 === 資訊科學與工程研究所 === 107 === In active tracking task, the agent needs to determine policies through the view from camera to keep the target object in its sight. In this thesis, we proposed a neural network model with fully-convolutional Siamese structure, which takes the target’s image a...
| Main Authors: | , |
|---|---|
| Other Authors: | |
| Format: | Others |
| Language: | en_US |
| Published: |
2018
|
| Online Access: | http://ndltd.ncl.edu.tw/handle/t8xx3y |