篇名 |
A Novel Online Teaching Effect Evaluation Model Based on Visual Question Answering
|
---|---|
並列篇名 | A Novel Online Teaching Effect Evaluation Model Based on Visual Question Answering |
作者 | Yanqing Cui、Guangjie Han、Hongbo Zhu |
英文摘要 | The paper proposes a novel visual question answering (VQA)-based online teaching effect evaluation model. Based on the text interaction between teacher and students, we give a guide-attention (GA) model to discover the directive clues. Combining the self-attention (SA) models, we reweight the vital feature to locate the critical information on the whiteboard and students’ faces and further recognize their content and facial expressions. Three branches of information are encoded into the feature vectors to be fed into a bidirectional GRU network. With the real labels of the students’ answers annotated by two teachers and the predicted labels from the text and facial expression feedback, we train the chained network. Experiment reports a couple of competitive performance in the 2-class and 5-class tasks on the self-collected dataset, respectively.
|
起訖頁 | 093-100 |
關鍵詞 | VQA、Facial expression recognition、Online teaching effect evaluation、GRU、Guide-attention |
刊名 | 網際網路技術學刊 |
期數 | 202201 (23:1期) |
出版單位 | 台灣學術網路管理委員會 |
DOI |
|
QR Code | |
該期刊 上一篇
| Dynamic Pyramid Attention Networks for multi-orientation object detection |
該期刊 下一篇
| Automatic Docking Optimization Method of Bed and Chair Based on Multi-sensor Information Fusion |