Haikyuu!!

If we go for somewhat safe here, it'll mean we never changed!Sugawara Koushi This quote summarizes the new Karasuno boy's volleyball team. Throughout all of the three seasons, change and risk have been the only constant for them. Haikyuu is the first sports Anime I have watched (although I have read some sports Manga before) … Continue reading Haikyuu!!

My First EDM Night!

The final event of Thomso 2018, Wargasm, a party filled with explosively loud music, high college students and electricity in the air, signed off the annual cultural fest of IIT Roorkee with a bass drop! So, this is my third Thomso after coming to IIT Roorkee and for various silly reasons my first time attending … Continue reading My First EDM Night!

Variational Autoencoder Explained

Variational encoders (VAEs) are generative models, in contrast to typical standard neural networks used for regression or classification tasks. VAEs have diverse applications from generating fake human faces and handwritten digits to producing purely "artificial" music. This post will explore what a VAE is, the intuition behind it and also the tough looking (but quite … Continue reading Variational Autoencoder Explained

A Look at Image Segmentation using CNNs

Image segmentation is the task in which we assign a label to pixels (all or some in the image) instead of just one label for the whole image. As a result, image segmentation is also categorized as a dense prediction task. Unlike detection using rectangular bounding boxes, segmentation provides pixel accurate locations of objects in … Continue reading A Look at Image Segmentation using CNNs

Paper Explanation: Net2Net – Accelerating Learning via Knowledge Transfer

Motivation One of the biggest challenges during designing new neural network architectures is time. During real-world workflows, one often trains many different neural networks during the experimentation and design process. This is a wasteful process in which each new model is trained from scratch. In a typical workflow, one trains multiple models, with each model … Continue reading Paper Explanation: Net2Net – Accelerating Learning via Knowledge Transfer

Paper Explanation: Binarized Neural Networks: Training Neural Networks with Weights and Activations Constrained to +1 or −1

Motivation What if you want to do some real-time face detection/recognition using a deep learning system running on a pair of glasses? What if you want your alarm clock to be able to record and analyze your sleep and conditions around you and come up with the most optimal way of waking you up each … Continue reading Paper Explanation: Binarized Neural Networks: Training Neural Networks with Weights and Activations Constrained to +1 or −1