Tweeted By @tanmingxing
Happy to introduce EfficientNetV2: Smaller Models and Faster Training
— Mingxing Tan (@tanmingxing) April 2, 2021
Achieved faster training and inference speed, AND also with better parameters efficiency.
Arxiv: https://t.co/YHWEb8pHmR
Thread 1/4 pic.twitter.com/LY2oZ4tSbN