Hao Tang

Interested in doing a PhD in machine learning at the University of Edinburgh? If so, check out the ML-Systems PhD Programme:
mlsystems.uk

Application deadline is 22nd January (next week).

The School of Informatics at the University of Edinburgh is hiring a lecturer or reader in embodied natural language processing. Apply by 31 Jan 2025 at edin.ac/4fqgawg

Back then Kaldi hadn't existed yet. HTK was still the dominant toolkit. I cannot remember if there was MLP implemented in HTK. I was mostly using a neural network toolkit (I think from ICSI) that just trained 3-layer MLPs. (Do people still have the source code? Somebody should put it on github.)

Hybrid (Bourlard and Morgan, 1989) and tandem (Hermansky et al., 2000) approaches were well established already. I think the tandem approach had an upper hand, until it was convincingly displaced in 2012. (But hey, we are using HuBERT features nowadays, not very different from a tandem approach.)

People were very skeptical, mainly because nobody knew where to even start to reproduce the results. (The results were finally reproduced years later in Kaldi.)

Feeling a bit nostalgic. I vividly remember, despite being in a gigantic venue (that was in Dallas), it was so so packed when the DBN results on phone recognition were presented in ICASSP 2010.