Skip to main content
Fig. 5 | Plant Methods

Fig. 5

From: Deep learning based high-throughput phenotyping of chalkiness in rice exposed to high night temperature

Fig. 5

Examples of Grad-CAM (ResNet-101) heatmaps for 10 sample chalky seed images (5 on the left side and 5 on the right side). For each seed, heatmaps corresponding to the following four layers are shown: (1) ResNet101 layer1_2_conv2; (2) ResNet101 layer2_0_conv2; (3) ResNet101 layer3_1_conv2; (4) ResNet101 layer4_1_conv3. The IoU values obtained for three thresholds T (20%, 40% and 60%, respectively) are shown under each heatmap

Back to article page