Title: StuArt: Individualized Classroom Observation of Students with Automatic Behavior Recognition and Tracking accepted by ICASSP2023
-
[Notes]
- I am very sorry to inform you that the dataset used cannot be made public due to the privacy issues of primary school students. Relevant researchers can only collect, annotate and process the data by themselves.
- The proposed StuArt has be integrated into our project AIClass, which is an Automatic Teaching Assistance System Towards Classrooms for K-12 Education. Please refer it for more details.
-
Figure 2
Top: examples of K-12 classroom scenarios. Bottom: some samples of annotated behaviors.
| A | B | C | D |
|---|---|---|---|
 |
 |
 |
 |
| Handraising | Standing | Sleeping | Yawning | Smiling |
|---|---|---|---|---|
 |
 |
 |
 |
 |
- Figure 3
Qualitative behavior detection results. Yellow, green, cyan, red, blue and black boxes are hand-raisings, standing, sleeping, yawning, smiling and teacher, respectively.
| 1 | 2 | 3 |
|---|---|---|
 |
 |
 |
- Figure 5
Left: visualizations of two non-interactive designs. The current data corresponds to the frame shown in Figure 4 top. Right: visualizations of two non-interactive designs.
| SeatTable | LinkList |
|---|---|
 |
 |
| ActivateHeatmap | PointFlow |
 |
 |