Current image quality (resolution, lighting, contrast, ...) allows to clearly identify occupation of a cavity starting from the second row onwards, as they increase in size. In the image, obviously a bee nested in the hole in row 2, column 5:
What is hard to recognize is that there are also nests in the first row, columns 2 and 3. The holes might be a little bit brighter to the sandy filling materials, but hard to recognize for the eye and probably any neural net as well.
A more close-up photograph with the smartphone (flashlight TURNED ON) reveals it:
So existing capabilities regarding camera settings need to be investigated for taking better images allowing for first row classification as well.
Current image quality (resolution, lighting, contrast, ...) allows to clearly identify occupation of a cavity starting from the second row onwards, as they increase in size. In the image, obviously a bee nested in the hole in row 2, column 5:
What is hard to recognize is that there are also nests in the first row, columns 2 and 3. The holes might be a little bit brighter to the sandy filling materials, but hard to recognize for the eye and probably any neural net as well.
A more close-up photograph with the smartphone (flashlight TURNED ON) reveals it:
So existing capabilities regarding camera settings need to be investigated for taking better images allowing for first row classification as well.