Herman Kamper

lecture_dtw_notebook. Jupyter Notebook

bayes_gmm. Bayesian Gaussian mixture models in Python.

data414. Data Analytics 414

vqwordseg. Unsupervised phone and word segmentation using dynamic programming on self-supervised VQ features.

linearvc. Voice conversion with just linear regression.

autoencoders_mnist. Different types of autoencoders illustrated on MNIST using TensorFlow.

speech_dtw. Dynamic time warping (DTW) functions for specifically speech alignment.

nlp817. Natural Language Processing 817

recipe_swbd_wordembeds. Shell

couscous. Siamese neural networks for representation learning using Theano.

eskmeans. Embedded segmental K-means (ES-KMeans) in Python.

speech_correspondence. Correspondence and autoencoder neural network training for speech using Pylearn2.

segmentalist. Unsupervised word segmentation and clustering of speech

globalphone_awe. Multilingual acoustic word embedding approaches applied and evaluated on GlobalPhone data.

suzerospeech2019. Stellenbosch University ZeroSpeech 2019 System

recipe_zs2017_track2. Recipe for applying the embedded segmental k-means model to the ZeroSpeech2017 Track 2 challenge.

recipe_bucktsong_awe_py3. Unsupervised acoustic word embeddings evaluated on Buckeye English and NCHLT Xitsonga data in Python 3.

bucktsong_segmentalist. Unsupervised segmentation and clustering of Buckeye English and NCHLT Xitsonga corpora.

yfacc. YFACC: A Yorùbá speech-image dataset for cross-lingual keyword localisation through visual grounding

semantic_flickraudio. A data set of semantic keyword spotting labels for the Flickr Audio Captions Corpus.

recipe_vision_speech_flickr. Using computer vision to ground unlabelled speech.

stellenbosch_ee_report_template. A LaTeX template for reports following the guidelines of the E&E department at Stellenbosch University.

recipe_bucktsong_awe. Unsupervised acoustic word embeddings evaluated on Buckeye English and NCHLT Xitsonga data in Python 2.7.

recipe_semantic_flickraudio. Semantic speech retrieval with a visually grounded model of untranscribed speech.

bucktsong_eskmeans. Unsupervised segmentation and clustering of the Buckeye English and NCHLT Xitsonga datasets using the ES-KMeans algorithm.

kamperh.github.io. SCSS

dpdp_aernn. Duration-penalized dynamic programming (DPDP) autoencoding recurrent neural network (AE-RNN) in Python.

ss414. Systems and Signals 414

flickr_semantic_qbe_eval. Evaluation code for semantic QbE on the Flickr8k Audio Captions Corpus

teaching_portfolio. Teaching Portfolio