Update README.md
Browse files
README.md
CHANGED
@@ -108,12 +108,12 @@ Test data consists of six Sentinel-2 mosaics:
|
|
108 |
|
109 |
|Location|Date|Number of annotations|
|
110 |
|--------|----|---------------------|
|
111 |
-
|Bothnian sea|2021-07-14|
|
112 |
-
|Bothnian sea|2022-06-24|
|
113 |
-
|Bothnian sea|2022-08-13|
|
114 |
-
|Kvarken|2022-06-17|
|
115 |
-
|Kvarken|2022-07-12|
|
116 |
-
|Kvarken|2022-08-26|
|
117 |
|
118 |
#### Factors
|
119 |
|
@@ -140,21 +140,21 @@ Precision and Recall with IoU-threshold of 0.5, mAP50 and mAP.
|
|
140 |
|
141 |
| Model | ('Precision', 'max') | ('Precision', 'mean') | ('Precision', 'min') | ('Recall', 'max') | ('Recall', 'mean') | ('Recall', 'min') | ('mAP', 'max') | ('mAP', 'mean') | ('mAP', 'min') | ('mAP50', 'max') | ('mAP50', 'mean') | ('mAP50', 'min') |
|
142 |
|:--------|-----------------------:|------------------------:|-----------------------:|--------------------:|---------------------:|--------------------:|-----------------:|------------------:|-----------------:|-------------------:|--------------------:|-------------------:|
|
143 |
-
| yolov8n |
|
144 |
-
| yolov8s |
|
145 |
-
| yolov8m |
|
146 |
-
| yolov8l |
|
147 |
-
| yolov8x |
|
148 |
|
149 |
Models with best performance for test set for each model architecture:
|
150 |
|
151 |
| Model | Fold | Precision | Recall | mAP50 | mAP |
|
152 |
|:------|------|-----------|--------|-------|-----|
|
153 |
-
|yolov8n| 1| 0.
|
154 |
-
|yolov8s| 4| 0.
|
155 |
-
|yolov8m| 4| 0.
|
156 |
-
|yolov8l| 1| 0.
|
157 |
-
|yolov8x| 1| 0.
|
158 |
|
159 |
### Compute Infrastructure
|
160 |
|
|
|
108 |
|
109 |
|Location|Date|Number of annotations|
|
110 |
|--------|----|---------------------|
|
111 |
+
|Bothnian sea|2021-07-14|450|
|
112 |
+
|Bothnian sea|2022-06-24|424|
|
113 |
+
|Bothnian sea|2022-08-13|399|
|
114 |
+
|Kvarken|2022-06-17|83|
|
115 |
+
|Kvarken|2022-07-12|183|
|
116 |
+
|Kvarken|2022-08-26|88|
|
117 |
|
118 |
#### Factors
|
119 |
|
|
|
140 |
|
141 |
| Model | ('Precision', 'max') | ('Precision', 'mean') | ('Precision', 'min') | ('Recall', 'max') | ('Recall', 'mean') | ('Recall', 'min') | ('mAP', 'max') | ('mAP', 'mean') | ('mAP', 'min') | ('mAP50', 'max') | ('mAP50', 'mean') | ('mAP50', 'min') |
|
142 |
|:--------|-----------------------:|------------------------:|-----------------------:|--------------------:|---------------------:|--------------------:|-----------------:|------------------:|-----------------:|-------------------:|--------------------:|-------------------:|
|
143 |
+
| yolov8n | 0.85001 | 0.840136 | 0.82782 | 0.82951 | 0.804012 | 0.78738 | 0.38816 | 0.380828 | 0.37637 | 0.84525 | 0.833424 | 0.81883 |
|
144 |
+
| yolov8s | 0.86717 | 0.854216 | 0.84347 | 0.84939 | 0.84065 | 0.83222 | 0.41098 | 0.406258 | 0.40374 | 0.86933 | 0.861404 | 0.84934 |
|
145 |
+
| yolov8m | 0.86108 | 0.853192 | 0.84191 | 0.87385 | 0.846722 | 0.83 | 0.41739 | 0.410742 | 0.40496 | 0.87772 | 0.862594 | 0.84602 |
|
146 |
+
| yolov8l | 0.86911 | 0.863254 | 0.85604 | 0.86468 | 0.841572 | 0.82725 | 0.41694 | 0.411712 | 0.40505 | 0.88288 | 0.867134 | 0.85743 |
|
147 |
+
| yolov8x | 0.86411 | 0.856008 | 0.85045 | 0.86086 | 0.845044 | 0.83029 | 0.42065 | 0.411532 | 0.40231 | 0.87069 | 0.863538 | 0.85316 |
|
148 |
|
149 |
Models with best performance for test set for each model architecture:
|
150 |
|
151 |
| Model | Fold | Precision | Recall | mAP50 | mAP |
|
152 |
|:------|------|-----------|--------|-------|-----|
|
153 |
+
|yolov8n| 1| 0.820204|0.838353| 0.764|0.285|
|
154 |
+
|yolov8s| 4| 0.843822|0.860479| 0.788|0.303|
|
155 |
+
|yolov8m| 4| 0.858263|0.874616| 0.814|0.324|
|
156 |
+
|yolov8l| 1| 0.840311|0.863553| 0.799|0.323|
|
157 |
+
|yolov8x| 1| 0.855134|0.859865| 0.795|0.316|
|
158 |
|
159 |
### Compute Infrastructure
|
160 |
|