3 ms·
Note: This was published in 2015, before / around the same time as other encoder-decoder-like segmentation architectures like SegNet or U-Net.
by zo7 10y ago
Note: This was published in 2015, before / around the same time as other encoder-decoder-like segmentation architectures like SegNet or U-Net.
- sweezyjeezy 10y agoYah was gona say, no one really uses pooling / unpooling in convnets any more, SOTA tends to be more about strided convolutions / unconvolutions. Not really sure why this is on the front page....