Zin / Steel Defect Classification Public

Target: Arduino Portenta H7 (Cortex-M7 480MHz) Clone this project

Welcome to Edge Impulse, the largest community of edge AI developers!
This is a public Edge Impulse project, use the navigation bar to see all data and models in this project; or clone to retrain or deploy to any edge device.

Steel Defect Classification

Images

About this project

This public Edge Impulse project does not have a README yet. Clone this project to add new data or retrain this project, or to deploy this project to a device.

Data (1,573 items)

View all

img_01_425008000_00572.jpg.2u4io807

ID: 83161250

Label: water_spot

img_06_3403333900_00758.jpg.2u4ikk6u

ID: 83161595

Label: punching_hole

img_02_425646200_00062.jpg.2u4ijocj

ID: 83161957

Label: oil_spot

img_03_425390000_00022.jpg.2u4ijoep

ID: 83161952

Label: oil_spot

img_02_3436813700_00329.jpg.2u4ioabp

ID: 83161164

Label: water_spot

img_01_425392100_00018.jpg.2u4ii30a

ID: 83162321

Label: crescent_gap

img_03_425389800_00017.jpg.2u4imt4d

ID: 83161481

Label: welding_line

img_03_3403393700_00942.jpg.2u4ikhnv

ID: 83161699

Label: punching_hole

Model accuracy (Unoptimized (float32))

Validation set

80.3%

Test set

On-device performance (Arduino Portenta H7 (Cortex-M7 480MHz), Quantized (int8))

Latency

610 ms.

Peak RAM usage

2.6M

Flash usage

2.9M

Run this model

On any device

Clone project

Dataset summary

Data collected

1,573 items

Labels

crescent_gap, inclusion, oil_spot, punching_hole, waist_folding, water_spot, welding_line

Project info

Project ID	87224
Project version	1
License	BSD 3-Clause Clear
No. of views	9,393
No. of clones	8