Deep Learning
In-depth understanding of Neural Networks and its working
Today using neural network architectures, we can achieve complex tasks like object recognition in images, automatic speech recognition (ASR), machine translation, image captioning, video classification, and a lot more that are challenging for machines to perform. It can do complex tasks using neural networks because artificial neural networks are the most powerful learning models in the field of machine learning. Neural networks can achieve arguably every task that the human brain can perform…
Seeing Through AI Eyes: The Technology Behind Style Transfer and Object Detection
In this blog, we'll dive deep into two computer vision applications that are changing how we interact with images, i.e., Style transfer and Object detection. We will explore the underlying working and technology used for these two applications. In recent months, social media has been transformed into a Ghibli wonderland, with users across platforms sharing photos of themselves and their surroundings reimagined with a Ghibli filter. We will explore how similar outputs can be obtained using…
How optimization in deep learning work?
Mathematical optimization is an extremely powerful field of mathematics that underpins much of what data scientists, implicitly, or explicitly, utilize regularly. Nearly all machine learning algorithms make use of optimization theory to obtain model convergence. Take, for example, a classification problem, we seek to minimize "log loss" by choosing the optimal parameters or weights of the model. In general, mathematical optimization can be thought of as a mechanism by which algorithms learn. A…
Seeing Through Layers: The Ultimate Guide to Convolutional Neural Networks
Convolutional Neural Networks (CNNs) are specialized neural networks that have revolutionized computer vision and image processing. While they share similarities with traditional Artificial Neural Networks (read post In-depth understanding of Neural Networks and its working to know about ANN), CNNs have unique architectural elements that make them particularly effective for visual data processing. Challenges in Image Processing Let's consider an object detection task to identify the vehicles…