← Authors

Martin Jansche

11 papers · 2001–2019 · 6 conferences · across top CS/AI conferences

Achievements

🐝 Cross-Pollinator (5) 🌈 Renaissance Researcher (6) 🐣 Hot Topic Early Bird 🧭 Keyword Pioneer πŸ—ΊοΈ Taxonomy Completionist (11) 🌍 Conference Polyglot (6) πŸƒ Academic Marathon (18) πŸŒ‰ Interdisciplinary Bridge 🌟 Keyword Trendsetter Combo (3) 🐺 Lone Wolf (5) 🌱 Topic Pioneer πŸ’Ž Century Club (11) πŸ“ˆ Trend Setter

Conferences

INTERSPEECH (3) ACL (2) EMNLP (2) NAACL (2) COLING (1) CONLL (1)

Top co-authors

Alexander Gutkin (2) Marelie Davel (1) Noemie Elhadad (1) Charl van Heerden (1) Lijun Feng (1) Steven Abney (1) Matt Huenerfauth (1) Linne Ha (1) Michael Riley (1) Oddur Kjartansson (1)

Keywords

binary classification (1) speech recognition (1) information extraction (1) speech processing (1) text analysis (1) readability assessment (1) feature comparison (1) voicemail transcription (1) speech transcription (1) spoken language understanding (1) connectionist temporal classification (1) sampling algorithm (1) linguistic feature (1) cross-lingual analysis (1) phonological feature (1) multilingual speech synthesis (1) path sampling (1) neural classifier (1) phoneme inventory (1) ctc decoding (1)

Papers

Sampling from Stochastic Finite Automata with Applications to CTC Decoding INTERSPEECH 2019
Cross-Lingual Consistency of Phonological Features: An Empirical Study INTERSPEECH 2019
Rapid Development of TTS Corpora for Four South African Languages INTERSPEECH 2017
A Comparison of Features for Automatic Readability Assessment COLING 2010
OpenFst: An Open-Source, Weighted Finite-State Transducer Library and its Applications to Speech and Language NAACL 2009
A Maximum Expected Utility Framework for Binary Sequence Labeling ACL 2007
Maximum Expected F-Measure Training of Logistic Regression Models EMNLP 2005
Parametric Models of Linguistic Count Data ACL 2003
Information Extraction from Voicemail Transcripts EMNLP 2002
Named Entity Extraction with Conditional Markov Models and Classifiers CONLL 2002
Re-Engineering Letter-to-Sound Rules NAACL 2001