Pushkar Khetrapal's repositories
SemanticBoxNet
The SemanticBoxNet is a Deep Learning Model which predicts semantic mask and bounding boxes in a single Network.
EfficientPS
EfficientPS is a Panoptic Segmentation architecture which includes Mask R-CNN, Semantic Segmentation and fusion of both outputs. EfficientPS was developed by Rohit Mohan, Abhinav Valada ( University of Freiburg, Germany ) in 2020.
Face-Recognition
Working on a CNN model which predicts the name of the student who enters the class. Maintaining a database of all the students with their time in and out.
3-Days-Machine-Learning-Workshop
A 3 day Machine Learning hands on workshop organised by TSG-GTBIT on 4th-5th-6th october 2019.
chatbot-automation-gmail
A Gmail voice recognition assistent which convert audio into text( subject + message) and automates the further proccess of sending the mail.
Classification-of-MEG-waves-in-EEG-and-optimized-with-TensorRT
Determining muscular electric signal in EEG brainwaves. And further optimized the with TensorRT (Multi-Input to output).
H-CNN
This project proposes a different architecture of Deep learning that takes ‘n’ different uncorrelated features of same image/signal parallelly into non-shareable convolution input layers in the same network to predict classes. That is referred to as 1/2-D Hybrid-Convolutional Neural Network.
Cognitive-Motor-Controller
A Brain Computer Interface Model which Classify EEG signals, evoked potential by brain activity with help of Convolutional 1D Neural Network (Tensorflow).
cuda_programming
Code from the "CUDA Crash Course" YouTube series by CoffeeBeforeArch
Mask_RCNN
Mask R-CNN for object detection and instance segmentation on Keras and TensorFlow
pcl
Point Cloud Library (PCL)
PyTorch-Simple-MaskRCNN
A PyTorch implementation of simple Mask R-CNN
realtime_panoptic
Official PyTorch implementation of CVPR 2020 Oral: Real-Time Panoptic Segmentation from Dense Detections
tensorrtx
Implementation of popular deep learning networks with TensorRT network definition API
vision
Datasets, Transforms and Models specific to Computer Vision
Yet-Another-EfficientDet-Pytorch
The pytorch re-implement of the official efficientdet with SOTA performance in real time and pretrained weights.