Improved Encoder-Decoder Temporal Action Detection Algorithm

Table 1. Parameters of feature extraction network
View table
Table 1. Parameters of feature extraction network
Block Kernel size Number of channels
Conv1 7×7 64
Conv2_x $[\begin{array}{l} 1 \times 1 \\ 3 \times 3 \\ 1 \times 1 \end{array}]$ ×3 $[\begin{array}{l} 64 \\ 64 \\ 256 \end{array}]$ ×3
Conv3_x $[\begin{array}{l} 1 \times 1 \\ 3 \times 3 \\ 1 \times 1 \end{array}]$ ×4 $[\begin{array}{l} 128 \\ 128 \\ 512 \end{array}]$ ×4
Conv4_x $[\begin{array}{l} 1 \times 1 \\ 3 \times 3 \\ 1 \times 1 \end{array}]$ ×6 $[\begin{array}{l} 256 \\ 256 \\ 1024 \end{array}]$ ×6
Conv5_x $[\begin{array}{l} 1 \times 1 \\ 3 \times 3 \\ 1 \times 1 \end{array}]$ ×3 $[\begin{array}{l} 512 \\ 512 \\ 2048 \end{array}]$ ×3

Table 2. Recognition accuracy rate of each action
View table
Table 2. Recognition accuracy rate of each action
Dataset Action Accuracy /%
MERL Shopping Reach to shelf 77.8
Retract from shelf 79.3
Hand in shelf 81.6
Inspect the product 80.4
Inspect the shelf 81.2

Table 3. Effectiveness of various module on the algorithm
View table
Table 3. Effectiveness of various module on the algorithm
Dataset VggNet16 ResNet50 ED-TCN Improved ED-TCN mAP /%
MERL Shopping √ √ 24.3
MERL Shopping √ √ 25.6
MERL Shopping √ √ 29.3
GTEA √ √ 25.8
GTEA √ √ 27.2
GTEA √ √ 30.2

Table 4. Seg-F1 of different algorithms on different datasets
View table
Table 4. Seg-F1 of different algorithms on different datasets
Dataset ED-TCN Improved ED-TCN Seg-F1@10 Seg-F1@25 Seg-F1@50
MERL Shopping √ 86.7 85.1 72.9
MERL Shopping √ 89.2 87.4 74.8
GTEA √ 72.2 69.3 56.0
GTEA √ 76.8 71.9 58.5

Table 5. Comparison of results of different algorithms on the MERL Shopping dataset
View table
Table 5. Comparison of results of different algorithms on the MERL Shopping dataset
Algorithm Accuracy /% mAP /% Seg-F1@10 Seg-F1@25 Seg-F1@50
MSN Det 64.6 29.5 46.4 42.6 25.6
MSN Seg 76.3 24.2 80.0 78.3 65.4
Dilated TCN 76.4 26.3 79.9 78.0 67.5
ED-TCN 79.0 25.5 86.7 85.1 72.9
Improved ED-TCN 82.4 29.3 89.2 87.4 74.8

Tools

Get Citation

Copy Citation Text

Yue Wang, Hansong Su, Gaohua Liu. Improved Encoder-Decoder Temporal Action Detection Algorithm[J]. Laser & Optoelectronics Progress, 2021, 58(20): 2020001

Download Citation

EndNote(RIS)BibTex Plain Text

Set citation alerts for article

Save article for my favorites