Victoria Ellison

Key Skills

Software

Labelbox

Top Subject Matter

No subject matter listed

Top Data Types

Image

Video

Audio

Top Label Types

Bounding Box

Point Key Point

Segmentation

Classification

Tracking

Emotion Recognition

Freelancer Overview

I am an experienced AI Specialist with three years dedicated to data annotation and quality assurance for machine learning projects. My background includes hands-on work with image classification, object detection, and audio transcription, using industry-standard tools such as Labelbox, CVAT, SuperAnnotate, and VGG Image Annotator. I have a strong track record of maintaining dataset accuracy above 98% through meticulous quality control and error tracking. My experience spans annotating images, videos, and audio—covering tasks like bounding boxes, segmentation, keypoints, and object tracking—while collaborating closely with AI engineers to refine labeling guidelines and ensure high-quality training data. I am passionate about leveraging my skills in dataset management, performance monitoring, and metadata organization to support the development of robust AI models.

Labeling Experience

High precison Multimodal data annotation for Computer vision model

LabelboxAudioBounding BoxPoint Key Point

Currently working on an audio data annotation project supporting speech recognition and NLP model training. Responsible for high-accuracy transcription of audio recordings, including speaker segmentation, timestamping, and intent classification. Performed emotion recognition labeling, background noise identification, and audio quality assessment to improve model robustness across diverse acoustic environments. Annotated multilingual and accented speech datasets while ensuring clarity and consistency. Maintained over 98% accuracy by following strict annotation guidelines, conducting quality assurance reviews, and correcting inconsistencies. Managed large datasets efficiently while meeting tight deadlines and adhering to data confidentiality standards. Contributed to improving AI model performance by identifying edge cases such as overlapping speech, low-volume recordings, heavy accents, and noisy backgrounds.

Currently working on an audio data annotation project supporting speech recognition and NLP model training. Responsible for high-accuracy transcription of audio recordings, including speaker segmentation, timestamping, and intent classification. Performed emotion recognition labeling, background noise identification, and audio quality assessment to improve model robustness across diverse acoustic environments. Annotated multilingual and accented speech datasets while ensuring clarity and consistency. Maintained over 98% accuracy by following strict annotation guidelines, conducting quality assurance reviews, and correcting inconsistencies. Managed large datasets efficiently while meeting tight deadlines and adhering to data confidentiality standards. Contributed to improving AI model performance by identifying edge cases such as overlapping speech, low-volume recordings, heavy accents, and noisy backgrounds.

2024

High Precision Multimodern data Annotation for computer vision models

LabelboxVideoBounding BoxPoint Key Point

Currently working on a large-scale video annotation project supporting computer vision model training. Responsible for frame-by-frame object detection using bounding boxes, polygon annotation, and multi-object tracking across dynamic scenes. Annotated moving objects including pedestrians, vehicles, animals, and environmental elements in real-world scenarios. Performed action recognition labeling and temporal segmentation to help models understand motion patterns and behavior sequences. Ensured high annotation accuracy (98%+) by strictly following labeling guidelines, conducting self-QA checks, and participating in peer reviews. Managed high-volume datasets while maintaining consistency in object IDs across frames for accurate tracking. Contributed to improving model performance by identifying edge cases such as occlusion, motion blur, low-light conditions, and overlapping objects.

Currently working on a large-scale video annotation project supporting computer vision model training. Responsible for frame-by-frame object detection using bounding boxes, polygon annotation, and multi-object tracking across dynamic scenes. Annotated moving objects including pedestrians, vehicles, animals, and environmental elements in real-world scenarios. Performed action recognition labeling and temporal segmentation to help models understand motion patterns and behavior sequences. Ensured high annotation accuracy (98%+) by strictly following labeling guidelines, conducting self-QA checks, and participating in peer reviews. Managed high-volume datasets while maintaining consistency in object IDs across frames for accurate tracking. Contributed to improving model performance by identifying edge cases such as occlusion, motion blur, low-light conditions, and overlapping objects.

2024

Computer Vision & Audio Data Annotation for AI Model Training

LabelboxImageBounding BoxPoint Key Point

Worked on large-scale AI training datasets involving image, video, and audio annotation. For computer vision projects, I performed object detection using bounding boxes, polygon annotation, semantic segmentation, and multi-object tracking across video frames. The datasets included urban street scenes, retail environments, and real-world object interactions. For audio projects, I completed speech transcription, speaker labeling, intent classification, and emotion tagging to support NLP and speech recognition models. Maintained over 98% annotation accuracy through strict QA processes, guideline adherence, cross-review validation, and error tracking. Managed datasets exceeding 50,000 annotated instances under tight deadlines while ensuring consistency and metadata organization.

Worked on large-scale AI training datasets involving image, video, and audio annotation. For computer vision projects, I performed object detection using bounding boxes, polygon annotation, semantic segmentation, and multi-object tracking across video frames. The datasets included urban street scenes, retail environments, and real-world object interactions. For audio projects, I completed speech transcription, speaker labeling, intent classification, and emotion tagging to support NLP and speech recognition models. Maintained over 98% annotation accuracy through strict QA processes, guideline adherence, cross-review validation, and error tracking. Managed datasets exceeding 50,000 annotated instances under tight deadlines while ensuring consistency and metadata organization.

2022 - 2024

Education

Work History