Ping Q.

Key Skills

Software

Other

Don't disclose

Top Subject Matter

Multimodal motion capture dataset concept

sensor + mobile vision data annotation pipeline.

Generative image workflow optimization and visual output evaluation.

Top Data Types

Image

Text

Document

Top Task Types

Data Collection

Prompt + Response Writing (SFT)

Freelancer Overview

Multimodal Motion Capture Dataset Concept (coarse labeling workflow planning). Core strengths include Other, ComfyUI, and Don't disclose. Education includes Bachelor of Finance, Fujian Normal University Concord College (2025). AI-training focus includes data types such as 3D Sensor, Image, and Text and labeling workflows including Data Collection, Evaluation, and Rating.

Labeling Experience

AI-Assisted Visual Workflow Exploration (prompt engineering + iterative evaluation)

Don't discloseTextPrompt + Response Writing (SFT)

The candidate created experimental AI-assisted visual systems for presentations, posters, and interface concepts with an emphasis on repeatable prompt-driven outputs. They built reusable prompting structures to preserve consistent visual styles across projects. Through iterative output refinement and prompt optimization, they effectively carried out evaluation-focused prompt engineering that can support future supervised training data creation. • Built reusable prompting structures for consistent visual outputs. • Conducted iterative refinement and prompt optimization loops. • Explored typography hierarchy, cinematic composition, and editorial layouts. • Used generative AI to generate and refine visual design candidates.

2026 - Present

ComfyUI Workflow Research & Optimization (iterative visual evaluation/consistency)

Image

The candidate conducted independent research on ComfyUI node-based image generation workflows with an emphasis on visual quality control. They explored style consistency and image compositing while optimizing the visual pipeline for more reliable outputs. The iterative refinement and troubleshooting supported consistent labeling of visual attributes for future evaluation. • Tested node-based image generation workflows in ComfyUI. • Explored style consistency and compositing strategies. • Investigated GPU memory optimization and pipeline troubleshooting. • Experimented with cinematic image generation and advanced visual prompting.

2026 - Present

Multimodal Motion Capture Dataset Concept (coarse labeling workflow planning)

Other3D SensorData Collection

The candidate designed a conceptual multimodal data collection pipeline for a future AI training dataset using wearable sensors and mobile vision. They researched MediaPipe/YOLO-based coarse labeling workflows and planned standardized JSON dataset structures for scalable annotation. The work emphasized preparing data for human-review pipelines and downstream training usage. • Planned multimodal dataset schema using JSON structures. • Researched MediaPipe and YOLO workflows for initial coarse labeling. • Investigated scalable data annotation and human-review processes. • Considered wearable sensor signals combined with mobile vision inputs.

2026 - Present

Education

Work History