Karthik Yearning Deep Learning

Statistically stopping of neural network training

Paper: Statistically Significant Stopping of Neural Network Training Github: Code Repository Much learning of neural network does not take place once the optimal values are found, the condition does not impact the final accuracy of the model. According to the runtime perspective, this is of great significance when numerous neural networ... Read more

Visualize feature maps

In this article, I will visualize the feature maps in a neural network. Since the focus of this article is to visualize the feature maps, I am using a tutorial neural network training script from PyTorch official website. This tutorial uses Transfer learning with Resnet50 architecture. The complete tutorial script can be found here. Visualizin... Read more

Exploring Siamese Network

Exploring Simple Siamese Representation Learning Siamese network is two parallel network sharing same weights to maximize the similarity between each network input. Siamese network is trained in an unsupervised manner. A image is augmented to create a positive pair for which the similarity score is ideally 1, on the other hand, a negativ... Read more

Multi-Model Ensemble via Adversarial Learning

Paper This paper presents a method for compressing large, complex trained ensembles into a single network, where knowledge from a variety of trained deep neural networks (DNNs) is distilled and transferred to a single DNN. The source knowledge networks are called teacher model, and the learning network is called student network. Adversari... Read more

Complete IoU

The loss functions are the major driving force in training a good model. In Object detection and Instance segmentation tasks, the most widely used loss function is Intersection over Union (IOU). In Enhancing Geometric Factors for Object Detection and Instance Segmentation Loss function. paper, a new loss function called as Complete Intersection ... Read more

Named Entity Recognition

Named Entity Recognition is the process of extracting entities present in sentences. This is an important aspect in information extraction from text documents. NER can also be used in conversational chats to extract entities such as location, time and many more. NER can also be trained to extract person names in information extraction of legal... Read more

Text Summarization Architectures

Text Summarizing is an important NLP Task. It comprises of two categories, abstractive summarization and extractive summarization. However, the real world application of text summarization is challenging due to human readability and quality. I will try to explain different research approaches and architectures employed. This blog post will foc... Read more

CenterNet

CenterNet: Keypoint Triplets for Object Detection This paper presents an efficient solution which explores the visual patterns within each cropped region with minimal costs. We build our framework upon a representative one-stage keypoint-based detector named CornerNet. Our approach, named CenterNet, detects each object as a triplet, rather ... Read more

Optimization in Deep Learning

Paper: An overview of gradient descent optimization algorithms This article will discuss about different optimization techniques that are widely employed in Neural Networks. This article focuses on optimizer selection strategy for desired Neural Networks. Gradient Descent Gradient descent is a way to minimize an objective function J(θ) pa... Read more