Dipeshtamboli/Image-Classification-and-Localization-using-Multiple-Instance-Learning

Stars
73
Rank 432,145 (Top 9 %)
Language
Python
Created over 4 years ago
Updated over 4 years ago

Dipeshtamboli/Image-Classification-and-Localization-using-Multiple-Instance-Learning

Dipeshtamboli

There are no reviews yet. Be the first to send feedback to the community and the maintainers!

Official code for Breast Cancer Histopathology Image Classification and Localization using Multiple Instance Learning

PosNegDM-Reinforced-Sequential-Decision-Making-for-Sepsis-Treatment

Official code for "Reinforced Sequential Decision-Making for Sepsis Treatment: The POSNEGDM Framework with Mortality Classifier and Transformer" published at JBHI.

Cross-Domain-FSL-via-NSAE

PyTorch implementation of Boosting the Generalization Capability in Cross-Domain Few-shot Learning via Noise-enhanced Supervised Autoencoder

Interactive-GradCAM

Easy to follow GradCAM visualization - Google collab notebooks where you just have to upload the image and mention the target class to get the feature visualization for models trained on COCO and Imagenet dataset.

Jupyter Notebook

multisource_person_reid

Person re-identification to observe domain shift. Models trained and tested on the self domain as well as on cross domains(i.e. Models trained on Market1501 and tested on DukeMTMC)

zsl_cgan

GPU-Benchmarking

A code to benchmark GPU performance on different models

emotion-detection-using-TF

A ready to use repository utilising the TensorFlow model for emotion detection. Can use pre-trained model directly for inferencing.

GAN_for_clothes

A simple demonstration of CycleGAN to change the colour of the clothes. Here generated image will have the same object content but different colour. Here domain A and B represent different cloth colours(i.e. Green and Maroon)

Autonated-Dispensary

Contains the codes for an Automated Dispensary system. The system builds a classifier for images of pills, powered by retraining the final layer of MobileNet architecture. It further contains implementations to localize the positions of certain objects in the omage through contour detection and a simple sliding window. Further, it has code to commnicate with an Arduino-UNO to receive orders and control stepper motors.