issues
search
OpenDroneMap
/
ODMSemantic3D
An open photogrammetry dataset of classified 3D point clouds for automated semantic segmentation. CC BY-SA 4.0
Other
15
stars
7
forks
source link
Added toledo dataset
#14
Closed
HeDo88TH
closed
1 year ago
u4gbot
commented
1 year ago
Comparison for brighton_beach_small (Differences)
Accuracy: 94.11% vs 96.10% | Label | Accuracy | Recall | Precision | F1 | | -------------------- | ---------- | ---------- | ---------- | ---------- | | | 66.95% | 95.07% | 69.36% | 0.80 | | low_vegetation | 32.20% | 40.42% | 61.27% | 0.49 | | | -34.75% | -54.65% | -8.09% | -0.31 | | | | | | | | | 45.35% | 83.98% | 49.64% | 0.62 | | human_made_object | 39.40% | 39.48% | 99.46% | 0.57 | | | -5.95% | -44.50% | 49.82% | -0.06 | | | | | | | | | 14.64% | 99.56% | 14.65% | 0.26 | | building | 20.68% | 44.58% | 27.84% | 0.34 | | | 6.04% | -54.98% | 13.19% | 0.09 | | | | | | | | | 94.06% | 94.12% | 99.93% | 0.97 | | ground | 96.34% | 98.68% | 97.59% | 0.98 | | | 2.28% | 4.57% | -2.34% | 0.01 | | | | | | |
Comparison for sheffield_park_small (Differences)
Accuracy: 97.85% vs 23.50% | Label | Accuracy | Recall | Precision | F1 | | -------------------- | ---------- | ---------- | ---------- | ---------- | | | 95.84% | 98.10% | 97.66% | 0.98 | | low_vegetation | 18.51% | 99.99% | 18.51% | 0.31 | | | -77.33% | 1.89% | -79.14% | -0.67 | | | | | | | | | 34.76% | 94.35% | 35.50% | 0.52 | | human_made_object | 0.71% | 0.74% | 16.12% | 0.01 | | | -34.05% | -93.61% | -19.39% | -0.50 | | | | | | | | | 92.46% | 94.00% | 98.27% | 0.96 | | building | 46.87% | 60.09% | 68.06% | 0.64 | | | -45.59% | -33.90% | -30.21% | -0.32 | | | | | | | | | 98.04% | 98.33% | 99.70% | 0.99 | | ground | 1.62% | 1.63% | 67.94% | 0.03 | | | -96.42% | -96.69% | -31.76% | -0.96 | | | | | | |
Average (2 datasets)
Accuracy: 95.98% vs 59.80%
Label
Accuracy
Recall
Precision
F1
81.40%
96.58%
83.51%
0.89
low_vegetation
25.36%
70.21%
39.89%
0.40
-56.04%
-26.38%
-43.62%
-0.49
40.06%
89.17%
42.57%
0.57
human_made_object
20.05%
20.11%
57.79%
0.29
-20.00%
-69.06%
15.22%
-0.28
53.55%
96.78%
56.46%
0.61
building
33.78%
52.33%
47.95%
0.49
-19.77%
-44.44%
-8.51%
-0.12
96.05%
96.22%
99.82%
0.98
ground
48.98%
50.16%
82.77%
0.51
-47.07%
-46.06%
-17.05%
-0.47
HeDo88TH
commented
1 year ago
!skip
Comparison for brighton_beach_small (Differences)
Accuracy: 94.11% vs 96.10% | Label | Accuracy | Recall | Precision | F1 | | -------------------- | ---------- | ---------- | ---------- | ---------- | | | 66.95% | 95.07% | 69.36% | 0.80 | | low_vegetation | 32.20% | 40.42% | 61.27% | 0.49 | | | -34.75% | -54.65% | -8.09% | -0.31 | | | | | | | | | 45.35% | 83.98% | 49.64% | 0.62 | | human_made_object | 39.40% | 39.48% | 99.46% | 0.57 | | | -5.95% | -44.50% | 49.82% | -0.06 | | | | | | | | | 14.64% | 99.56% | 14.65% | 0.26 | | building | 20.68% | 44.58% | 27.84% | 0.34 | | | 6.04% | -54.98% | 13.19% | 0.09 | | | | | | | | | 94.06% | 94.12% | 99.93% | 0.97 | | ground | 96.34% | 98.68% | 97.59% | 0.98 | | | 2.28% | 4.57% | -2.34% | 0.01 | | | | | | |Comparison for sheffield_park_small (Differences)
Accuracy: 97.85% vs 23.50% | Label | Accuracy | Recall | Precision | F1 | | -------------------- | ---------- | ---------- | ---------- | ---------- | | | 95.84% | 98.10% | 97.66% | 0.98 | | low_vegetation | 18.51% | 99.99% | 18.51% | 0.31 | | | -77.33% | 1.89% | -79.14% | -0.67 | | | | | | | | | 34.76% | 94.35% | 35.50% | 0.52 | | human_made_object | 0.71% | 0.74% | 16.12% | 0.01 | | | -34.05% | -93.61% | -19.39% | -0.50 | | | | | | | | | 92.46% | 94.00% | 98.27% | 0.96 | | building | 46.87% | 60.09% | 68.06% | 0.64 | | | -45.59% | -33.90% | -30.21% | -0.32 | | | | | | | | | 98.04% | 98.33% | 99.70% | 0.99 | | ground | 1.62% | 1.63% | 67.94% | 0.03 | | | -96.42% | -96.69% | -31.76% | -0.96 | | | | | | |Average (2 datasets)
Accuracy: 95.98% vs 59.80%