SHeaP: Self-Supervised Head Geometry Predictor Learned via 2D Gaussians
New paper SHeaP is out: please see here !
nlml
nlml: thoughts on machine learning
Latest posts
Semi-Supervised Learning (and more): Kaggle Freesound Audio Tagging
An overview of semi-supervised learning and other techniques I applied to a recent Kaggle competition.
Getting Champion Coordinates from the LoL Minimap using Deep Learning
Using a GAN and a ConvLSTM to go from minimap from to champion coordinates: This post was originally published on Medium.
In Raw Numpy: t-SNE
This is the first post in the In Raw Numpy series. This series is an attempt to provide readers (and myself) with an understanding of some of the most frequently-used machine learning methods by going through the math and intuition, and implementing it using just python and numpy.
Using Tensorboard Embeddings Visualiser with Numpy Arrays
Tensorboard’s embeddings visualiser is great. You can use it to visualise and explore any set of high dimensional vectors (say, the activations of a hidden layer of a neural net) in a lower-dimensional space.
Adversarial Neural Cryptography in Theano
Last week I read Abadi and Andersen’s recent paper [1], Learning to Protect Communications with Adversarial Neural Cryptography. I thought the idea seemed pretty cool and that it wouldn’t be too tricky to implement, and would also serve as an ideal project to learn a bit more Theano. This post describes the paper, my implementation, and the results.
Detecting Music BPM using Neural Networks - Update
This post is a brief update to my previous post about using a neural network to detect the beats per minute (BPM) in short sections of audio.
Detecting Music BPM using Neural Networks
I have always wondered whether it would be possible to detect the tempo (or beats per minute, or BPM) of a piece of music using a neural network-based approach. After a small experiment a while back, I decided to make a more serious second attempt. Here’s how it went.
Facebook Recruiting IV
The ‘Facebook Recruiting IV: Human or bot?’ competition just ended on Kaggle. For those unfamiliar with the competition, participants downloaded a table of about 7 million bids, which corresponded to another table of around 6,000 bidders. For 4,000 of those bidders, you had to estimate the probability that they were a human or bot, based on the remaining 2,000 bidders, whose bot status was given.