I fine-tuned an Inception V3 model provided in AWS SageMaker to detect COVID-19 Rapid Test Results (see the image below for an example). I provided about 20 pictures of negative and about 20 pictures of positive tests for the training. All pictures were taken with slightly changing angles and positions. However, when testing the fined-tuned model, the recognition did not work at all.

Is the deviation between both image classes is too small (only changing red bars). Is there any other way a machine learning model can detect such small deviations between a positive and negative test without heavy image pre-processing?

COVID-19 Rapid Test Results

  • $\begingroup$ It will be easier if you clip the read portion and make it a simple array and apply a simple rule-based logic on pixels values. $\endgroup$
    – 10xAI
    Aug 27 at 4:37

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.