Deep Learning

Get to know the concept of Deep Learning while you wait for your eggs to boil. — As a subfield of Machine Learning, Deep Learning maps input data(x/features) to output(y/targets). The journey of mapping out is carried out by layers which act to transform the input data, numbers of layers could vary from tens to tens of hundred depending on what we are trying to map out…

Showcasing the power of deep learning models used in time series when compared to previous machine learning models — Time series forecasting can occur when scientific predictions are based on historical time-stamped data. This involves building models through historical analysis and using them to make observations and drive future strategic decision-making. We live in a day of age where energy conservation and usage of renewable energy sources is a…

Image segmentation is the process of partitioning an image into regions or objects which helps in the meaningful representation of the… — This is a small research project implemented as part of the Computer Vision course(CSCI-B657) by professor David Crandall offered at Indiana University Bloomington. This project can essentially be divided into two topics, Semantic Segmentation, and Attention mechanism.

Quantization and fusion for faster inference — Modern deep learning research is mainly focused on creating and improving new, better, and optimized solutions for various problems such as object detection, segmentation, self-supervised learning, etc. This is helping many businesses and startups start using more accessible AI solutions for their products. However, when it comes to real-world applications…

In the world of digital marketing, Ads are bought and sold programmatically through auctions RTB (Real-time bidding) which means that every… — In the world of digital marketing, Ads are bought and sold programmatically through auctions RTB (Real-time bidding) which means that every time there is an ad inventory that is available from SSPs(Supply-Side Platforms)/Ad Networks, multiple DSPs(Demand-Side Platforms) on behalf of advertisers bid for advertising at the same spot. Whoever wins…

Above are three pairs of graphs. The top graph is the original function, f(x), and the bottom graph is the derivative, f’(x). — Above are three pairs of graphs. The top graph is the original function, f(x), and the bottom graph is the derivative, f’(x). What do you notice about each pair? If the slope of f(x) is negative, then the graph of f’(x) will be below the x-axis. If the slope of…

S⁴L: Combines Self-Supervised Approach and Semi-Supervised Approach — S⁴L: Self-Supervised Semi-Supervised Learning S⁴L, by Google Research, Brain Team 2019 ICCV, Over 400 Citations (Sik-Ho Tsang @ Medium) Self-Supervised Learning, Semi-Supervised Learning, Image Classification By unifying self-supervised learning and semi-supervised learning, the framework of self-supervised semi-supervised learning (S⁴L) is proposed.