Tool Presence Detection on Cholec80 (test)
95.6AP (Bipolar)R + C + CL
Evaluation results
| Method | Links | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| R + C + CLModel=R + C + CL2018.12 | 95.6 | 97.5 | 99.7 | 99.8 | 74.7 | 86.9 | 96.1 | 92.9 | |
| R + CL + CModel=R + CL + C2018.12 | 95.6 | 97.1 | 99.8 | 99.9 | 77.4 | 76.1 | 93.9 | 91.4 | |
| R + CLModel=R + CL2018.12 | 93.8 | 97.5 | 99.5 | 99.9 | 65.1 | 90.3 | 74 | 88.5 | |
| R + CM1_maskModel=R + CM1_mask2018.12 | 92.6 | 96.9 | 99.8 | 99.8 | 60.9 | 85.1 | 78.6 | 87.7 | |
| R + CM1Model=R + CM12018.12 | 91.9 | 80.3 | 96.7 | 99.4 | 85.2 | 50.6 | 88.3 | 84.6 | |
| R + CM4_maskModel=R + CM4_mask2018.12 | 90.9 | 88.5 | 99.6 | 99.8 | 66.2 | 48.5 | 91 | 83.6 | |
| MTRCNet-CLMulti-task learning=true2019.07 | 90.1 | 89.8 | 84.7 | 95.6 | 88.2 | 86.7 | 88.9 | 89.1 | |
| R + CM4Model=R + CM42018.12 | 89.4 | 85.4 | 95.9 | 99.5 | 89.5 | 69.3 | 87.1 | 87.9 | |
| EndoNetbackbone=AlexNet, training=transfer learning, framework=multi-task2016.02 | 86.9 | 80.1 | 84.8 | 95.6 | 74.4 | 58.6 | 86.8 | 81 | |
| EndoNetMulti-task learning=true2019.07 | 86.9 | 80.1 | 84.8 | 95.6 | 74.4 | 58.6 | 86.8 | 81 | |
| ToolNetbackbone=AlexNet, training=transfer learning2016.02 | 85.9 | 79.8 | 84.7 | 95.5 | 73 | 60.9 | 86.3 | 80.9 | |
| ToolNetMulti-task learning=false2019.07 | 85.9 | 79.8 | 84.7 | 95.5 | 73 | 60.9 | 86.3 | 80.9 | |
| DPM2016.02 | 60.6 | 68.4 | 82.3 | 93.4 | 40.5 | 23.4 | 40 | 58.4 | |
| DPMMulti-task learning=false2019.07 | 60.6 | 68.4 | 82.3 | 93.4 | 40.5 | 23.4 | 40 | 58.4 |