Nuo C.

Key Skills

Software

Label Studio

LabelImg

Doccano

Top Subject Matter

Artificial Intelligence / Machine Learning

Data Annotation & Labeling

NLP / Natural Language Processing

Top Data Types

Video

Image

Text

Top Task Types

Computer Programming/Coding

Question Answering

Text Summarization

Classification

Evaluation/Rating

Object Detection

Freelancer Overview

I have solid experience in AI data processing, data labeling and model training data production. I am proficient in text sorting, content evaluation, Q&A scoring, text rewriting and various AI data annotation tasks. Skilled in basic data cleaning, information sorting and content quality checking. Familiar with standard operating processes of remote AI training work, able to arrange working time reasonably, finish assigned tasks efficiently and strictly follow task rules to deliver high-quality labeled data. I am careful, patient and quick to learn, able to steadily complete long-term stable remote part-time jobs.

Labeling Experience

Research on High-Risk Special Operations and Gas Leakage Risk Monitoring and Early Warning Technology

3D SensorClassification

Created an anomaly detection and early warning pipeline for high-risk special operations and gas leakage monitoring using multi-sensor time-series data. Processed multiple sensors’ time-series streams to extract statistical features that reduce computation while preserving relevant signal behavior. Compared multiple sequence/time-series models (MLP, XGBoost, SVM, LSTM) and selected a Pulse-feature plus XGBoost architecture for real-time hazard detection. • Implemented multi-source time-series processing and time-domain statistical feature extraction. • Built and benchmarked anomaly detection models, ultimately adopting Pulse features + XGBoost. • Designed real-time early warning system logic to meet seconds-level response and a ≤20s response-time constraint. • Achieved >95% anomaly detection accuracy with false positive rate <5% and delivered completed prototypes.

2024 - 2025

Design of an electronic nose for lung cancer screening based on exhaled breath detection

Diagnosis

Developed an electronic-nose style lung cancer screening approach based on exhaled breath electronic sensing signals for non-invasive diagnosis. Performed sensor signal preprocessing including response conversion and FFT-based frequency selection to downsample the effective frequency representation. Built a 1D-CNN and GRU hybrid model with a fusion attention mechanism to capture local sensor features and temporal/spatial dependencies for binary classification. • Used 1D-CNN for feature extraction from segmented time slices while preserving partial temporal information. • Combined attention-derived features with GRU hidden states and attention scores. • Produced binary classification probabilities via a final fully connected layer. • Achieved ~90% accuracy and ~90.91% sensitivity compared with multiple baselines.

2024 - 2025

Design of a Wearable Vital Sign Monitoring and Early Warning System (signal processing and algorithm design)

Diagnosis

Built an ML-based wearable vital sign monitoring and early warning system using sensor-derived signals for abnormal detection. Implemented ECG and PPG preprocessing (EEMD + stationary wavelet transform) and respiratory filtering to prepare clean time-series inputs. Trained a convolutional neural network for fall detection and evaluated heart/respiratory/SpO2-related analytics for real-world readiness. • Used tri-axial acceleration data to classify normal activities vs falls. • Applied AMPD peak detection for respiratory rate calculation and thresholding for heart rate estimation. • Performed signal preprocessing to remove baseline drift and high-frequency noise. • Benchmarked fall detection accuracy to exceed 93%.

2024 - 2024

Education

Work History