Micah Cross

Key Skills

Software

Appen

Data Annotation Tech

Remotasks

Scale AI

Top Subject Matter

No subject matter listed

Top Data Types

Image

Text

Video

Top Task Types

Audio Recording

Classification

Data Collection

Question Answering

Text Summarization

Freelancer Overview

I have extensive experience supporting data labeling and AI training processes across various modalities, including text, image, audio, and video data. My role involves curating high-quality datasets by ensuring accurate annotations aligned with project goals, such as sentiment analysis, object detection, named entity recognition, and speech transcription. I follow annotation guidelines meticulously, maintain consistency across large datasets, and flag ambiguous or edge cases for review to uphold data quality and integrity. Additionally, I contribute to refining labeling workflows by identifying inefficiencies and suggesting improvements, ensuring faster turnaround without sacrificing accuracy. My understanding of machine learning models allows me to anticipate how labeling decisions affect model performance, making my work both technically informed and contextually relevant. This collaborative and detail-oriented approach supports the development of robust, bias-aware AI systems.

Labeling Experience

Video

AppenVideoSegmentation

The project involved video segmentation for data annotation to support the training of computer vision models in tasks such as object tracking, action recognition, and scene understanding. Annotators were responsible for segmenting objects or regions of interest frame-by-frame within video sequences, using specialized annotation tools to draw precise boundaries and ensure temporal consistency across frames.

The project involved video segmentation for data annotation to support the training of computer vision models in tasks such as object tracking, action recognition, and scene understanding. Annotators were responsible for segmenting objects or regions of interest frame-by-frame within video sequences, using specialized annotation tools to draw precise boundaries and ensure temporal consistency across frames.

2024

Images

Data Annotation TechImageQuestion Answering

This project focused on image-based question answering for the purpose of training and validating computer vision and visual question answering (VQA) models. Annotators were tasked with examining images and providing accurate answers to related questions, which ranged from identifying objects and scenes to interpreting actions, relationships, or counting items. The goal was to generate high-quality annotated data to enhance AI systems’ ability to understand visual content and respond to natural language queries.

This project focused on image-based question answering for the purpose of training and validating computer vision and visual question answering (VQA) models. Annotators were tasked with examining images and providing accurate answers to related questions, which ranged from identifying objects and scenes to interpreting actions, relationships, or counting items. The goal was to generate high-quality annotated data to enhance AI systems’ ability to understand visual content and respond to natural language queries.

2023

Audio

Data Annotation TechAudioText Generation

This project involved audio-to-text generation for data annotation to support the development of automatic speech recognition (ASR) and natural language processing (NLP) systems. Annotators were tasked with listening to audio recordings—ranging from conversations and interviews to voice commands—and accurately transcribing spoken content into text. The primary goal was to produce high-quality, time-aligned transcripts that reflect true speech content, including punctuation, speaker turns, and non-verbal cues when applicable.

This project involved audio-to-text generation for data annotation to support the development of automatic speech recognition (ASR) and natural language processing (NLP) systems. Annotators were tasked with listening to audio recordings—ranging from conversations and interviews to voice commands—and accurately transcribing spoken content into text. The primary goal was to produce high-quality, time-aligned transcripts that reflect true speech content, including punctuation, speaker turns, and non-verbal cues when applicable.

2023 - 2024

Text

Data Annotation TechTextQuestion Answering

The project involved annotating textual data by reading passages and accurately answering corresponding questions to support the development of machine learning models for natural language understanding (NLU). Each task required careful comprehension of the text and selection or generation of contextually correct answers, ensuring alignment with predefined annotation guidelines. The goal was to create high-quality question-answer pairs that could be used to train and evaluate AI systems in areas such as reading comprehension, question answering, and information retrieval.

The project involved annotating textual data by reading passages and accurately answering corresponding questions to support the development of machine learning models for natural language understanding (NLU). Each task required careful comprehension of the text and selection or generation of contextually correct answers, ensuring alignment with predefined annotation guidelines. The goal was to create high-quality question-answer pairs that could be used to train and evaluate AI systems in areas such as reading comprehension, question answering, and information retrieval.

2023 - 2024

Education

Work History