Karthik Yearning Deep Learning

AI @ fb

From Satellite Imagery to Disaster Insights Large-Scale Visual Relationship Understanding Spatially Invariant Unsupervised Object Detection with Convolutional Neural Networks SGD Implicitly Regularizes Generalization Error Rethinking floating point for deep learning A2 -Nets: Double Attention Networks Explore-Exploit:... Read more

All about Seq2seq

Sequence to Sequence Learning Sequence to sequence model comprises of two sequence networks one at the encoder side and other at the decoder side. These sequence networks are RNN. The encoder networks maps the input sequence to a vector of a fixed dimension. The decoder network decodes the target sequence from the encoded vector. Seq2seq... Read more

Understanding CNN

This is a summarization of the Explanatory Graphs for CNNs paper. A graphical model is employed to understand the hierarchial feature representation hidden inside the CNN network which facilitates the classification , localization applications in Deep learning. The feature representation pattern from each filter is extracted to understand using... Read more

Gated Graph Sequence Neural Networks

Gated Graph Sequence Neural Networks Authors: Yujia Li & Richard Zemel Department of Computer Science, University of Toronto Toronto, Canada {yujiali,zemel}@cs.toronto.edu Marc Brockschmidt & Daniel Tarlow Microsoft Research Cambridge, UK {mabrocks,dtarlow}@microsoft.com ... Read more

Multi Target Prediction

In this blog, lets talk about multi target variables. In a generic machine learning model prediction scenario, where we will have input features and single target variable. Input Feature Output Feature x1 , x2 , x3 , x4… y Assigning multiple target variables would throw an error on line... Read more

Neural Network without Math

Let’s cover the fundamentals. What is Convolution? Convolution uses kernels/filter to extract information from images. Different kernels can extract different representations from images. For different tasks such as sharpening , edge detection , blurring different kernels can be employed. Kernel is matrix which slides upon the image to extrac... Read more

Summary - SGDR

SGDR: STOCHASTIC GRADIENT DESCENT WITH WARM RESTARTS Restart techniques are common in gradient-free optimization to deal with multimodal functions. Partial warm restarts are also gaining popularity in gradientbased optimization to improve the rate of convergence in accelerated gradient schemes to deal with ill-conditioned functions. In ... Read more

AI @ Cornell

Research Group Machine Learning Robotics Vision Projects Vision Recognition, Image Matching and Image Databases Object Recognition using the Hausdorff Distance Spatially Coherent Matching and Bayesian Recognition Flexible Object Recognition Motion, Stereo and Segmentation Image Segmentation using Local Variation Ma... Read more

List of AI Research Projects at Stanford.MIT.CMU

Stanford Intelligent wearable Robotics Safe feedback interactions in human autonomous vehicle systems Detailed understanding of human actions and behavior for smart vehicles Human centric autonomous and assistive driving Understanding driver state in laboratory and naturalistic environments ... Read more