Feedback for 2.3: The Dataset

Clarity: 4/5
Quality: 5/5
Relevance: 5/5

Errors:

-Around 6:50, you mention 16 images, but only 9 are shown. And because they are shuffled, it is not immediately a good example of augmentation, because it takes some (unnecessary) time to identify pictures that belong together.

Suggestions:

– Explanation of segmentation could be better. You explain everything orally, and only then show the examples. For someone who has never heard of segmentation, the examples come too late, because he/she will not have understood the oral explanation.

– It’s a bit of a pity that case 1 and 2 are so similar. That’s not necessarily a bad thing, because it also allows people to draw parallels, but we would certainly need an additional, strongly different case (not steel related). Case 3 alone will not be enough.

Location: AI for Materials Industry » Case 2 » 2.3: The Dataset

Clarity: 4/5
Quality: 5/5
Relevance: 5/5

Errors:

-Around 6:50, you mention 16 images, but only 9 are shown. And because they are shuffled, it is not immediately a good example of augmentation, because it takes some (unnecessary) time to identify pictures that belong together.

Suggestions:

– Explanation of segmentation could be better. You explain everything orally, and only then show the examples. For someone who has never heard of segmentation, the examples come too late, because he/she will not have understood the oral explanation.

– It’s a bit of a pity that case 1 and 2 are so similar. That’s not necessarily a bad thing, because it also allows people to draw parallels, but we would certainly need an additional, strongly different case (not steel related). Case 3 alone will not be enough.

  • Actions:

Leave a Reply

Your email address will not be published. Required fields are marked *