Ziang Zhou

About Me

My name is Ziang Zhou. In December 2023, I received M.Sc. of Intelligent Information Systems (MIIS) at Language Technology Institute (LTI), School of Computer Science (SCS) of Carnegie Mellon University. In May 2022, I received my dual B.Sc. of Data Science from Duke University and Duke Kunshan University.

Industry Experience

Currently I am working as a ML engineer at Pinterest.

In the summer of 2023, I interned at Pinterest Labs with the Ads Measurement Modeling Team as an Machine Learning Engineer. I worked on the self-supervised pretraining for User Match Prediction on Ads conversions.

Back in the summer of 2021, I interned at ByteDance AI Lab as an algorithm engineer, focusing on Dialect Machine Translation from Mandarin to Cantonese for the CapCut App.

Projects

• espnet/espnet: Contributed Automatic Speech Recognition (ASR) x (SLU) Spoken Language Understanding Recipe for Multimodal EmotionLines Dataset (MELD). [ recipe | 🤗 model card ]
• XenoPy: Python wrapper for Xeno-canto API 2.0. Supports multiprocessing. Supports installation through pip and conda.
• easybird: Python toolkit for Bird Activity Detection (BAD).
• Large Scale Birds Sound Recognition: A pipeline for Large Scale Birds Sound Detection and Recognition tasks.
• EmoTag: End-to-end emotion detection for Chinese audios.
• canvas-pilot: Command Line Interface (CLI) for Canvas Academic Platform.
• CMU Charger Hub: Decentralized charger sharing platform for Carnegie Mellon University students built on Ethereum blockchain.
• mit-han-lab/lite-transformer: Supported issues and contributed PR.
• GREMA: CLI tool to memorize GRE vocabulary.

Publications

• CCIS 2022: Ziang Zhou, Yanze Xu, Ming Li. “Detecting Escalation Level from Speech with Transfer Learning and Acoustic-Linguistic Information Fusion.”
• ICPR 2022: Yueran Pan, Jiaxin Wu, Ran Ju, Ziang Zhou, Jiayue Gu, Songtian Zeng, Lynn Yuan, Ming Li. “A Multimodal Framework for Automated Teaching Quality Assessment of One-To-Many Online Instruction Videos.”

R&D

During my Graduate studies, I work with Dr. Shinji Watanabe on Compositional ASR and TTS systems, and speech to speech systems.

During my undergraduate studies, I worked as a research assistant in Speech and Multimodal Intelligent Information Processing (SMIIP) Lab, with wonderful Dr. Ming Li. My primary focus is computational paralinguistics, including conversation escalation detection, speech emotion recognition, bird sound detection and recognition, etc.

I developed and maintained several python libraries such as XenoPy and easybird. My current research and engineering interests are mainly in computational linguistics. In my spare time, I also explore AI-based quantitative analysis, and general AI.

About this site

This is a personal blog site that shares both technologies and personal stuff, which may cover life records, reflections, important moments and decisions in my life.