gremlin97/RemoteSensingCorpus · Datasets at Hugging Face

text stringlengths 0 820
deviation of the performance.
L7 Irish L8 Biome
Backbone Pre-training Accuracy mIoU Accuracy mIoU
ResNet-18ImageNet 64.08 ±3.40 47.21 ±3.71 41.86 ±0.46 24.67 ±0.37
MoCo 74.79±2.20 59.77 ±2.79 42.70 ±5.02 27.33 ±4.14
SimCLR 34.80 ±11.36 21.46 ±8.53 39.17 ±5.12 24.44 ±3.89
ResNet-50ImageNet 61.77 ±3.27 44.75 ±3.41 45.73 ±6.08 29.78 ±5.23
MoCo 69.62±1.94 53.42 ±2.29 45.95±5.17 29.23 ±4.44
SimCLR 49.37 ±12.86 33.41 ±11.20 48.77±6.42 32.41 ±5.56
ViT-S16ImageNet 68.22 ±1.39 51.78 ±1.59 47.29±1.88 30.98 ±1.61
MoCo 86.65±0.43 76.03 ±0.67 46.66±3.59 30.33 ±3.14
SimCLR 82.65 ±0.27 70.47 ±0.35 42.33 ±1.80 26.99 ±1.67
ignore it. The fact that our SSL techniques work at all, let alone outperform ImageNet, demonstrates
the generalizability of our pre-trained model weights to different downstream applications.
Performance metrics for our land cover/land use tasks are reported in Table 2. Again, MoCo con-
sistently outperforms ImageNet in 25 out of 30 experiments across all sensors and product levels,
while SimCLR is unable to beat ImageNet in 24 out of 30 experiments. Performance gains by MoCo
are more modest in this task with a larger number of classes, but still reach as high as 6.60% overall
accuracy and 7.13 mIoU. There are exceptions to this, particularly for ETM+ TOA, but with ad-
ditional hyperparameter tuning of the pre-trained model it may be possible to exceed performance
of ImageNet. We attempted to use weights based on class frequency in our cross-entropy loss, but
these resulted in reduced accuracy and mIoU.
Figure 4 shows an example prediction made by a ResNet-18 backbone pre-trained on SSL4EO-L
using MoCo and a U-Net fine-tuned on CDL. Although the model is unable to predict detailed fea-
tures like roads and field corners, it removes much of the noise introduced by the pixel-wise decision
tree classifier used to produce CDL. Black pixels in the mask represent uncommon crop types that
are mapped to the background class. The model tends to pick the most common agricultural classes
like corn and soybean given no examples of these crop types in the training dataset. Although winter
wheat and fallow (idle farmland) are sometimes misclassified by the model, this is not unexpected.
Pasture Winter Wheat Fallow Corn Soybean Other
(a) Image
(b) Mask
(c) Prediction
Figure 4: Landsat 8 OLI SR image, ground truth mask, and prediction made by a U-Net with a
ResNet-18 backbone pre-trained using MoCo and SSL4EO-L and fine-tuned on CDL 2019.
8
Table 2: SSL4EO-L benchmark results. Overall accuracy and mean intersection over union (mIoU)
are reported for the test splits of the NLCD and CDL datasets for a range of sensors, product levels,
backbones, and pre-training techniques. All predictions are made by U-Nets with frozen backbones.
Satellite
(Sensor)Level
(Product)NLCD CDL
Backbone Pre-training Accuracy mIoU Accuracy mIoU
Landsats 4–5
(TM)Level-1
(TOA)ResNet-18ImageNet 65.63 48.84 66.11 49.38
MoCo 67.65 51.11 68.70 52.32
SimCLR 60.86 43.74 61.94 44.86
ResNet-50ImageNet 66.63 49.96 67.42 50.85
MoCo 68.75 53.28 69.45 53.20
SimCLR 62.05 44.98 62.80 45.77
ViT-S16ImageNet 68.93 52.59 68.27 51.83
MoCo 67.17 50.57 67.60 51.07
SimCLR 66.82 50.17 66.92 50.28
Landsat 7
(ETM+)Level-1
(TOA)ResNet-18ImageNet 66.11 49.38 65.84 49.08
MoCo 65.22 48.39 62.84 45.81
SimCLR 58.76 41.60 56.47 39.34
ResNet-50ImageNet 64.01 47.06 66.23 49.51
MoCo 66.60 49.92 64.12 47.19
SimCLR 57.17 40.02 54.95 37.88
ViT-S16ImageNet 62.06 45.01 57.67 40.52
MoCo 63.75 46.79 60.88 43.70
SimCLR 63.33 46.34 59.06 41.91
Level-2
(SR)ResNet-18ImageNet 63.34 46.34 60.70 43.58
MoCo 64.18 47.25 67.30 50.71
SimCLR 57.26 40.11 54.42 37.48
ResNet-50ImageNet 64.29 47.38 61.66 44.57
MoCo 64.37 47.46 62.35 45.30
SimCLR 57.79 40.64 55.69 38.59
ViT-S16ImageNet 63.54 46.56 51.38 34.57
MoCo 64.09 47.21 52.37 35.48
SimCLR 63.99 47.05 53.17 36.21
Landsats 8–9
(OLI/TIRS)Level-1
(TOA)ResNet-18ImageNet 66.40 49.70 65.21 48.38
MoCo 67.82 51.30 65.74 48.96
SimCLR 62.14 45.08 60.01 42.86
ResNet-50ImageNet 67.73 51.20 66.45 49.76
MoCo 69.17 52.87 67.29 50.70
SimCLR 64.66 47.78 62.08 45.01
ViT-S16ImageNet 65.52 48.72 62.38 45.33
MoCo 67.11 50.49 64.62 47.73
SimCLR 66.12 49.39 63.88 46.94
Level-2
(SR)ResNet-18ImageNet 65.46 48.65 62.88 45.85
MoCo 67.01 50.39 68.05 51.57
SimCLR 59.93 42.79 57.44 40.30
ResNet-50ImageNet 66.29 49.58 64.17 47.24
MoCo 67.44 50.88 65.96 49.21
SimCLR 63.65 46.68 60.01 43.17
ViT-S16ImageNet 65.71 48.93 62.78 45.75
MoCo 66.81 50.16 64.17 47.24
SimCLR 65.04 48.20 62.61 45.46
9
Winter wheat is planted in the fall and may be harvested before our summer imagery is taken. Sim-
ilarly, fallow can look much like pasture as weeds begin to grow in empty fields.
Fill Cloud Shadow Clear Thin Cloud Cloud

deviation of the performance.

L7 Irish L8 Biome

Backbone Pre-training Accuracy mIoU Accuracy mIoU

ResNet-18ImageNet 64.08 ±3.40 47.21 ±3.71 41.86 ±0.46 24.67 ±0.37

MoCo 74.79±2.20 59.77 ±2.79 42.70 ±5.02 27.33 ±4.14

SimCLR 34.80 ±11.36 21.46 ±8.53 39.17 ±5.12 24.44 ±3.89

ResNet-50ImageNet 61.77 ±3.27 44.75 ±3.41 45.73 ±6.08 29.78 ±5.23

MoCo 69.62±1.94 53.42 ±2.29 45.95±5.17 29.23 ±4.44

SimCLR 49.37 ±12.86 33.41 ±11.20 48.77±6.42 32.41 ±5.56

ViT-S16ImageNet 68.22 ±1.39 51.78 ±1.59 47.29±1.88 30.98 ±1.61

MoCo 86.65±0.43 76.03 ±0.67 46.66±3.59 30.33 ±3.14

SimCLR 82.65 ±0.27 70.47 ±0.35 42.33 ±1.80 26.99 ±1.67

ignore it. The fact that our SSL techniques work at all, let alone outperform ImageNet, demonstrates

the generalizability of our pre-trained model weights to different downstream applications.

Performance metrics for our land cover/land use tasks are reported in Table 2. Again, MoCo con-

sistently outperforms ImageNet in 25 out of 30 experiments across all sensors and product levels,

while SimCLR is unable to beat ImageNet in 24 out of 30 experiments. Performance gains by MoCo

are more modest in this task with a larger number of classes, but still reach as high as 6.60% overall

accuracy and 7.13 mIoU. There are exceptions to this, particularly for ETM+ TOA, but with ad-

ditional hyperparameter tuning of the pre-trained model it may be possible to exceed performance

of ImageNet. We attempted to use weights based on class frequency in our cross-entropy loss, but

these resulted in reduced accuracy and mIoU.

Figure 4 shows an example prediction made by a ResNet-18 backbone pre-trained on SSL4EO-L

using MoCo and a U-Net fine-tuned on CDL. Although the model is unable to predict detailed fea-

tures like roads and field corners, it removes much of the noise introduced by the pixel-wise decision

tree classifier used to produce CDL. Black pixels in the mask represent uncommon crop types that

are mapped to the background class. The model tends to pick the most common agricultural classes

like corn and soybean given no examples of these crop types in the training dataset. Although winter

wheat and fallow (idle farmland) are sometimes misclassified by the model, this is not unexpected.

Pasture Winter Wheat Fallow Corn Soybean Other

(a) Image

(b) Mask

(c) Prediction

Figure 4: Landsat 8 OLI SR image, ground truth mask, and prediction made by a U-Net with a

ResNet-18 backbone pre-trained using MoCo and SSL4EO-L and fine-tuned on CDL 2019.

8

Table 2: SSL4EO-L benchmark results. Overall accuracy and mean intersection over union (mIoU)

are reported for the test splits of the NLCD and CDL datasets for a range of sensors, product levels,

backbones, and pre-training techniques. All predictions are made by U-Nets with frozen backbones.

Satellite

(Sensor)Level

(Product)NLCD CDL

Backbone Pre-training Accuracy mIoU Accuracy mIoU

Landsats 4–5

(TM)Level-1

(TOA)ResNet-18ImageNet 65.63 48.84 66.11 49.38

MoCo 67.65 51.11 68.70 52.32

SimCLR 60.86 43.74 61.94 44.86

ResNet-50ImageNet 66.63 49.96 67.42 50.85

MoCo 68.75 53.28 69.45 53.20

SimCLR 62.05 44.98 62.80 45.77

ViT-S16ImageNet 68.93 52.59 68.27 51.83

MoCo 67.17 50.57 67.60 51.07

SimCLR 66.82 50.17 66.92 50.28

Landsat 7

(ETM+)Level-1

(TOA)ResNet-18ImageNet 66.11 49.38 65.84 49.08

MoCo 65.22 48.39 62.84 45.81

SimCLR 58.76 41.60 56.47 39.34

ResNet-50ImageNet 64.01 47.06 66.23 49.51

MoCo 66.60 49.92 64.12 47.19

SimCLR 57.17 40.02 54.95 37.88

ViT-S16ImageNet 62.06 45.01 57.67 40.52

MoCo 63.75 46.79 60.88 43.70

SimCLR 63.33 46.34 59.06 41.91

Level-2

(SR)ResNet-18ImageNet 63.34 46.34 60.70 43.58

MoCo 64.18 47.25 67.30 50.71

SimCLR 57.26 40.11 54.42 37.48

ResNet-50ImageNet 64.29 47.38 61.66 44.57

MoCo 64.37 47.46 62.35 45.30

SimCLR 57.79 40.64 55.69 38.59

ViT-S16ImageNet 63.54 46.56 51.38 34.57

MoCo 64.09 47.21 52.37 35.48

SimCLR 63.99 47.05 53.17 36.21

Landsats 8–9

(OLI/TIRS)Level-1

(TOA)ResNet-18ImageNet 66.40 49.70 65.21 48.38

MoCo 67.82 51.30 65.74 48.96

SimCLR 62.14 45.08 60.01 42.86

ResNet-50ImageNet 67.73 51.20 66.45 49.76

MoCo 69.17 52.87 67.29 50.70

SimCLR 64.66 47.78 62.08 45.01

ViT-S16ImageNet 65.52 48.72 62.38 45.33

MoCo 67.11 50.49 64.62 47.73

SimCLR 66.12 49.39 63.88 46.94

Level-2

(SR)ResNet-18ImageNet 65.46 48.65 62.88 45.85

MoCo 67.01 50.39 68.05 51.57

SimCLR 59.93 42.79 57.44 40.30

ResNet-50ImageNet 66.29 49.58 64.17 47.24

MoCo 67.44 50.88 65.96 49.21

SimCLR 63.65 46.68 60.01 43.17

ViT-S16ImageNet 65.71 48.93 62.78 45.75

MoCo 66.81 50.16 64.17 47.24

SimCLR 65.04 48.20 62.61 45.46

9

Winter wheat is planted in the fall and may be harvested before our summer imagery is taken. Sim-

ilarly, fallow can look much like pasture as weeds begin to grow in empty fields.

Fill Cloud Shadow Clear Thin Cloud Cloud