Prakash JayPatchnPack — Processing images using transformers of any shape & size.I was once having a conversation with one of my friend on why everything is square patches generation when most of the images/videos in…Feb 232Feb 232
Prakash JayMasked AutoEncodersMasked AutoEncoder is a simple algorithm to learn representations of an image in self-supervised way.Feb 121Feb 121
Prakash JayViTDet — The go to architecture for image foundation modelsViTDet, as of Jan 2024 is the go to architecture for all the vision tasks. It is used in segment-anything & ViTAE-Transformer has SOTA on…Feb 5Feb 5
Prakash JayUnderstanding AutoRegressive Image Models (AIM) — Introduced by Apple.Understanding Casual Masking in attention & PrefixLMJan 20Jan 20
Prakash JayCutout- Dropout in input space — Albumentations implementationCutout employs the strategy of randomly removing multiple small patches from an image, as opposed to a single patch, thereby causing the…Jun 24, 20231Jun 24, 20231
Prakash JayImage Classification Architectures reviewThis blog post gives a brief overview of the Image classification Architectures evolved since AlexNet till SENet.Jul 19, 20181Jul 19, 20181
Prakash JayThe intuition behind RetinaNetThe end goal of this blog post is to make readers intuitively understand the deep working of RetinaNet.Mar 23, 201826Mar 23, 201826
Prakash JayUsing Focal Loss for Deep Recommender systems.This blog post explains the approach I have taken during the 1 day hackathon hosted by Analytics Vidya. I stood 11th rank on public leader…Mar 15, 20182Mar 15, 20182
Prakash JayUnderstanding and Implementing Architectures of ResNet and ResNeXt for state-of-the-art Image…In this part-2/2 of blog post we will explore the optimal functions used in skip-connections of ResNet blocks. Discuss the ResNeXt…Feb 12, 20184Feb 12, 20184