minority4u
Software Developer in the field of Computer Vision and Deep Learning.
https://github.com/Cardio-AIHeidelberg
Pinned Repositories
cmr-domain-adaptation
Unsupervised Domain Adaptation (AX2SAX) for CMR images, Koehler et al. 2021, IEEE TMI
cmr-landmark-detection
Semantic Segmentation of CMR with a U-Net based architecture. Implemented in TF2.X. Trainings, prediction and evaluation scripts/notebooks for heatmap based right ventricle insertion point detection on cine CMR images. Koehler et al. 2022, BVM
3d-mri-domain-adaptation
Unsupervised Domain Adaptation (AX2SAX) for CMR images, Koehler et al. 2021, IEEE TMI
blockchain-ethereum-app
Ethereum based voting and smart contract implementation with test data and plug and play mining nodes and setup files.
cmr-landmark-detection
Semantic Segmentation of CMR with a U-Net based architecture. Implemented in TF2.X. Trainings, prediction and evaluation scripts/notebooks for heatmap based right ventricle insertion point detection on cine CMR images. Koehler et al. 2022, BVM
cmr-phase-detection
Self-supervised method for cardiac phase detection in 4D CMR. Model consists of a deformable registration part, the derivation of the per-voxel deformation angle and a rule-set based on the physiological properties of a contracting ventricle. Implemented in TF2.X. Koehler et al. 2022, STACOM workshop @ MICCAI
coockiecutter-ds-mod
Modified template for data-science, machine learning and deep learning projects.
data-science-file-walker
File-wrapper for unknown data exploration. Example implementations of different file types (wave, JSON etc.) to filter and analyse unknown data structures for data science projects.
functional-unet-tf
Functional U-Net for 2D/3D segmentation based on tf.keras
mobile-voting-app-android
Voting app for urban improvements. Supports different source formats to easily create new suggestions.
minority4u's Repositories
minority4u/blockchain-ethereum-app
Ethereum based voting and smart contract implementation with test data and plug and play mining nodes and setup files.
minority4u/coockiecutter-ds-mod
Modified template for data-science, machine learning and deep learning projects.
minority4u/mobile-voting-app-android
Voting app for urban improvements. Supports different source formats to easily create new suggestions.
minority4u/3d-mri-domain-adaptation
Unsupervised Domain Adaptation (AX2SAX) for CMR images, Koehler et al. 2021, IEEE TMI
minority4u/cmr-landmark-detection
Semantic Segmentation of CMR with a U-Net based architecture. Implemented in TF2.X. Trainings, prediction and evaluation scripts/notebooks for heatmap based right ventricle insertion point detection on cine CMR images. Koehler et al. 2022, BVM
minority4u/cmr-phase-detection
Self-supervised method for cardiac phase detection in 4D CMR. Model consists of a deformable registration part, the derivation of the per-voxel deformation angle and a rule-set based on the physiological properties of a contracting ventricle. Implemented in TF2.X. Koehler et al. 2022, STACOM workshop @ MICCAI
minority4u/data-science-file-walker
File-wrapper for unknown data exploration. Example implementations of different file types (wave, JSON etc.) to filter and analyse unknown data structures for data science projects.
minority4u/functional-unet-tf
Functional U-Net for 2D/3D segmentation based on tf.keras
minority4u/image-classification-tf
Multiclass image classification for technical drawings.
minority4u/machine-learning-algs
Implementation of different ML, DL and RL algorythms for educational purpose and as template for further work.
minority4u/rest-api-flask-orm
REST-APi with different routes, different db backbones, orm and docker/ cloud foundry deployment files.
minority4u/sentiment-analysis-tf-sl
Text tokenizer and preprocessing wrappers and different sentiment classifier pipelines (SVM, naive bayes, word2vec) based on scikit-learn, nltk and tf.
minority4u/virtual-machine
Implementation of an own stack, virtual machine, profiler, assembler, and example opcode to run in this machine.
minority4u/wav-file-trimmer
Stand-alone scripts to preprocess/trim wav files according to an rolling average.