A Method For Facial Pain Intensity Estimation Under Partial Nose and Mouth Occlusion
碩士 === 國立臺南大學 === 資訊工程學系碩士班 === 107 === Automatic pain intensity estimation system has a high value in clinical diagnosis. Since it is able to assist nursing personnel by performing immediate or long-term monitoring of multiple patients at the same time and also could be used as the basis for diagno...
| Main Authors: | , |
|---|---|
| Other Authors: | |
| Format: | Others |
| Language: | zh-TW |
| Published: |
2019
|
| Online Access: | http://ndltd.ncl.edu.tw/handle/p8675z |
| Summary: | 碩士 === 國立臺南大學 === 資訊工程學系碩士班 === 107 === Automatic pain intensity estimation system has a high value in clinical diagnosis. Since it is able to assist nursing personnel by performing immediate or long-term monitoring of multiple patients at the same time and also could be used as the basis for diagnosis. Most of current state-of-art techniques in pain assessment aim to estimate pain intensity under the facial images with no mask. But facial masked by medical supplies can be seen everywhere in medical environment. Our proposed method could estimate the pain intensity under partial occlusion images correctly, and it calls ABDN(Attention Based DenseNet) which is integrated the Attention Module and DenseNet . Extensive experiments show that our proposed model achieves a promising performance.
|
|---|