Yung Bing Cheng

Key Skills

Software

Appen

Clickworker

Google Cloud Vertex AI

OneForma

Telus

Other

Top Subject Matter

No subject matter listed

Top Data Types

Audio

Image

Text

Top Task Types

Classification

Data Collection

Evaluation Rating

Mapping

Translation Localization

Freelancer Overview

With over five years of experience as a Remote Data Annotator/Assistant at Neevo by DefinedCrowd, I have developed a strong expertise in AI training data, specializing in audio and image annotation. My key role involved recording and validating Cantonese and Mandarin Chinese audio samples to enhance speech recognition model training, ensuring high accuracy in pronunciation and natural speech patterns. This hands-on experience has honed my linguistic precision and attention to detail, making me adept at handling diverse audio datasets for AI applications. Additionally, I have contributed significantly to image-related AI projects by reviewing and correcting metadata for image tagging, as well as collecting and verifying locale-specific images and texts. This work improved the performance of AI classification models through consistent naming, categorization, and metadata accuracy. My background in managing multilingual audio data and visual content, combined with my ability to work remotely and adapt to various tools, sets me apart as a versatile and reliable candidate for AI training roles.

Labeling Experience

Validate the Pronunciation

OtherAudioEmotion Recognition

I contributed to a project titled "Validate the Pronunciation," where I assessed audio samples generated by an AI chatbot system, focusing on Cantonese and Mandarin. Using custom labeling software (Other), I evaluated the validity and naturalness of the audio, identifying issues such as speaking too fast, excessive background noise, unnatural pauses, and inaccurate pronunciation. My role involved providing detailed reasons for inaccuracies, ensuring high-quality emotion recognition annotations. This experience highlights my expertise in audio data labeling and my ability to enhance AI training for natural language processing in multilingual contexts.

I contributed to a project titled "Validate the Pronunciation," where I assessed audio samples generated by an AI chatbot system, focusing on Cantonese and Mandarin. Using custom labeling software (Other), I evaluated the validity and naturalness of the audio, identifying issues such as speaking too fast, excessive background noise, unnatural pauses, and inaccurate pronunciation. My role involved providing detailed reasons for inaccuracies, ensuring high-quality emotion recognition annotations. This experience highlights my expertise in audio data labeling and my ability to enhance AI training for natural language processing in multilingual contexts.

2025 - 2025

Map Evaluations

OneformaGeospatial Tiled ImageryClassificationTranslation Localization

I participated in a map evaluation project where I followed detailed instructions and assessments to analyze geospatial/tiled imagery. My role involved identifying and classifying user queries from sample data, determining their validity and types, such as classification, translation, localization, and mapping. After classifying the queries, I verified the accuracy of the requested locations, providing reasoned explanations for any discrepancies or mistakes. This project, conducted using One Forma software, showcased my expertise in LLM-related geospatial data labeling and my ability to ensure high-quality annotations for AI training.

I participated in a map evaluation project where I followed detailed instructions and assessments to analyze geospatial/tiled imagery. My role involved identifying and classifying user queries from sample data, determining their validity and types, such as classification, translation, localization, and mapping. After classifying the queries, I verified the accuracy of the requested locations, providing reasoned explanations for any discrepancies or mistakes. This project, conducted using One Forma software, showcased my expertise in LLM-related geospatial data labeling and my ability to ensure high-quality annotations for AI training.

2025 - 2025

Education

Work History