🚧 | Road crack segmentation using PyTorch

PyTorch extensions for fast R&D prototyping and Kaggle farming

Classification of Ionosphere dataset using pytorch neural networks.

Alternative loss function of binary cross entropy and focal loss

Tried to implement focal loss on EKG anomaly identification and Classification to address class imbalance as a challenge in this project

KL severity grading using SE-ResNet and SE-DenseNet architectures trained with Cross Entropy loss and Focal Loss. The hyperparameters of focal loss have been fine-tuned as well. Further, Grad-CAM has been implemented for visualization purposes.

Constrained self-supervised method with temporal ensembling for fiber bundle detection on anatomic tracing data

YOLO Series

PyTorch implementation of focal loss for multi-class semantic segmentation

Imbalanced classification with scikit-learn and PyTorch Lightning.

HistoSeg is an Encoder-Decoder DCNN which utilizes the novel Quick Attention Modules and Multi Loss function to generate segmentation masks from histopathological images with greater accuracy. This repo contains the code to Test and Train the HistoSeg

The project focuses on tackling challenges such as imbalanced data and skewed features. Through exploratory data analysis,and model training.The use of innovative techniques like focal loss and controlled oversampling allows us to address the imbalanced nature of the data and achieve better model performance.

Pytorch implementation of Class Balanced Loss based on Effective number of Samples

Adversarial Focal Loss: Asking Your Discriminator for Hard Examples.

CSE465 Pattern Recognition and Neural Network Project

Chinese NER(Named Entity Recognition) using BERT(Softmax, CRF, Span)

Classification for SMS Spam Collection Dataset using BERT

PyTorch implementation of focal loss for dense object detection

SSD-based object and text detection with Keras, SSD, DSOD, TextBoxes, SegLink, TextBoxes++, CRNN

This project aim to understad if a deep learning model is calibrated (average accuracy match average confident) using Reliability Diagram and perform a re-calibration by the training with Focal Loss.