3 ms·
We already did this for Object/Face recognition, it works but it's not the way to go. It's the way to go only if you don't have enough compute power (and data,
by gekoxyz 1y ago
We already did this for Object/Face recognition, it works but it's not the way to go. It's the way to go only if you don't have enough compute power (and data, I suspect) for a E2E network
- sixo 1y agoNo, it's what you do if your model architecture is capped out on its ability to profit from further training. Hand-wrapping a bunch of sub-models stands in for models that can learn that kind of substructure directly.