Small Components Inspection - Model testing - Edge Impulse

Mukesh Sankhla / Small Components Inspection Public

Target: Arduino UNO Q (Qualcomm QRB2210) Clone this project

This lists all test data. You can manage this data through Data acquisition.

Test data

Set the 'expected outcome' for each sample to the desired outcome to automatically score the impulse.

Sample name	Expected outcome	Precision score	Result
WIN_20260929_20_06_18_Pro	C5N, L2N, R5N	66%		Show classification Show in data acquisition Show in feature explorer
WIN_20260929_20_06_12_Pro	C1N, C2N, C3N, C4N, C5N, C6N, C7N, L1N, L2N, R1N, R3N, R4N, R5N	92%		Show classification Show in data acquisition Show in feature explorer
WIN_20260928_09_51_32_Pro	C1N, C2N, C3N, C4N, C5N, C6N, C7N, L1N, L2N, R1N, R3N, R4N, R5N	92%		Show classification Show in data acquisition Show in feature explorer
WIN_20260928_09_51_20_Pro	C1N, C2N, C3N, C4N, C5N, C6N, C7N, L1N, L2N, R1N, R3N, R4N, R5N	100%		Show classification Show in data acquisition Show in feature explorer
WIN_20260928_09_48_02_Pro (2)	C1A, C2A, C3A, C4A, C5A, C6A, C7A, L1A, L2A, R1A, R3A, R4A, R5A	84%		Show classification Show in data acquisition Show in feature explorer
WIN_20260928_09_47_58_Pro	C1A, C2A, C3A, C4A, C5A, C6A, C7A, L1A, L2A, R1A, R3A, R4A, R5A	100%		Show classification Show in data acquisition Show in feature explorer
WIN_20260928_09_47_01_Pro	C1A, C2A, C3A, C4A, C5A, C6A, C7A, L1A, L2A, R1A, R3A, R4A, R5A	92%		Show classification Show in data acquisition Show in feature explorer
WIN_20260928_09_46_50_Pro	C1A, C2A, C3A, C4A, C5A, C6A, C7A, L1A, L2A, R1A, R3A, R4A, R5A	76%		Show classification Show in data acquisition Show in feature explorer
WIN_20260928_09_46_46_Pro	C1A, C2A, C3A, C4A, C5A, C6A, C7A, L1A, L2A, R1A, R3A, R4A, R5A	84%		Show classification Show in data acquisition Show in feature explorer
WIN_20260928_09_43_59_Pro	C1A, C2A, C3A, C4A, C5A, C6A, C7A, L1A, L2A, R1A, R3A, R4A, R5A	92%		Show classification Show in data acquisition Show in feature explorer

Model testing output

Results

mAP@50

0.89

Metrics for Pretrained learn

Metric	Value
mAP	0.59
mAP@[IoU=50]	0.89
mAP@[IoU=75]	0.72
mAP@[area=small]	-1.00
mAP@[area=medium]	0.59
mAP@[area=large]	-1.00
Recall@[max_detections=1]	0.62
Recall@[max_detections=10]	0.62
Recall@[max_detections=100]	0.62
Recall@[area=small]	-1.00
Recall@[area=medium]	0.62
Recall@[area=large]	-1.00
Precision score (legacy)	90.0%

Feature explorer

Loading...

These users will be notified over email when this job finishes.

	Name
	Mukesh Sankhla