-
-
PyTorch Groups parameter
Dissecting Group Convolutions for RegNet models
Grouped Convolutions in Convolutional Neural Networks [Read More] -
An Image is Worth 16x16 Words
How the Transformer networks replace CNN for image classification
-
Attention is All You Need: An Overview
An in-depth assessment of the underlying mechanisms of Transformers
Avid computer vision readers know that lately, transformer models have become all the hype as they show a convenient way of training deep learning models without the sequential models like RNNs or making use of the convoluted ways of CNNs (pun-intended). As I treaded down this path, I realized there... [Read More] -
PseudoLiDAR and 4D Object Detection
First post in 4D object detection with PseudoLiDAR Input Series
PseudoLiDAR Approach [Read More]