New tutorials on http://keras.io : FixRes, a technique for boosting the performance of vision models by addressing the fact that they tend to be trained on crops of different sizes compared to the images they see at inference time.https://keras.io/examples/vision/fixres/ …
-
Show this thread
-
Created by
@RisingSayak. Check out the original paper:https://arxiv.org/abs/1906.064231 reply 1 retweet 10 likesShow this thread
Anything you can do to monitor and reduce the training/serving skew can make a big difference. The pitfalls of local generalization... :)
0 replies
1 retweet
7 likes
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.