111 citations · 462 across the 65 of their papers we have counts for
10 papers · 1 filter
Critical initialisation for deep signal propagation in noisy rectifier neural networks
Arnu Pretorius, Elan Van Biljon, Steve Kroon +1
Stochastic regularisation is an important weapon in the arsenal of a deep learning practitioner. However, despite recent theoretical advances, our understanding of how noise influe…
Feature exploration for almost zero-resource ASR-free keyword spotting using a multilingual bottleneck extractor and correspondence autoencoders
Raghav Menon, Herman Kamper, Ewald van der Westhuizen +2
We compare features for dynamic time warping (DTW) when used to bootstrap keyword spotting (KWS) in an almost zero-resource setting. Such quickly-deployable systems aim to support…
Multilingual and Unsupervised Subword Modeling for Zero-Resource Languages
Enno Hermann, Herman Kamper, Sharon Goldwater
Subword modeling for zero-resource languages aims to learn low-level representations of speech audio without using transcriptions or other resources from the target language (such…
Multimodal One-Shot Learning of Speech and Images
Ryan Eloff, Herman A. Engelbrecht, Herman Kamper
Imagine a robot is shown new concepts visually together with spoken tags, e.g. "milk", "eggs", "butter". After seeing one paired audio-visual example per class, it is shown a new s…
Truly unsupervised acoustic word embeddings using weak top-down constraints in encoder-decoder models
Herman Kamper
We investigate unsupervised models that can map a variable-duration speech segment to a fixed-dimensional representation. In settings where unlabelled speech is the only available…
Pre-training on high-resource speech recognition improves low-resource speech-to-text translation
Sameer Bansal, Herman Kamper, Karen Livescu +2
We present a simple approach to improve direct speech-to-text translation (ST) when the source language is low-resource: we pre-train the model on a high-resource automatic speech…
Learning Dynamics of Linear Denoising Autoencoders
Arnu Pretorius, Steve Kroon, Herman Kamper
Denoising autoencoders (DAEs) have proven useful for unsupervised representation learning, but a thorough theoretical understanding is still lacking of how the input noise influenc…
Fast ASR-free and almost zero-resource keyword spotting using DTW and CNNs for humanitarian monitoring
Raghav Menon, Herman Kamper, John Quinn +1
We use dynamic time warping (DTW) as supervision for training a convolutional neural network (CNN) based keyword spotting system using a small set of spoken isolated keywords. The…
Visually grounded cross-lingual keyword spotting in speech
Herman Kamper, Michael Roth
Recent work considered how images paired with speech can be used as supervision for building speech systems when transcriptions are not available. We ask whether visual grounding c…
Low-Resource Speech-to-Text Translation
Sameer Bansal, Herman Kamper, Karen Livescu +2
Speech-to-text translation has many potential applications for low-resource languages, but the typical approach of cascading speech recognition with machine translation is often im…