孙浩

Key Skills

Software

Label Studio

Roboflow

Top Subject Matter

Computer Vision, Object Detection & Image Annotation

UAV Remote Sensing & Geospatial Data

Intelligent Inspection, Defect Detection & AI Model Evaluation

Top Data Types

Image

Top Task Types

Object Detection

Freelancer Overview

First-year Master of Engineering student in Big Data Technology and Engineering at Tianjin Normal University, with hands-on project experience in AI-assisted remote sensing, UAV data processing, GIS visualization, and computer vision model integration. I have participated in building AI analysis modules for spectral inspection and UAV remote-sensing platforms, involving YOLO-based object detection, model weight configuration, inference API development, and front-end/back-end integration. My AI-related experience focuses on real-world detection scenarios such as forest fire hotspot detection, road crack detection, human-and-vehicle recognition, and photovoltaic panel defect detection. I am familiar with preparing task-specific detection workflows, validating model availability, troubleshooting missing model weights, and connecting AI inference results with visualization platforms and data-processing systems.

Labeling Experience

AI Analysis Module for Spectral Intelligent Inspection Platform

Object Detection

Worked on integrating real YOLO-based object detection into an AI analysis module for spectral/inspection tasks using UAV/remote-sensing related inputs. Configured model-weight paths, added model availability synchronization, and implemented a unified detection API that returns inference errors for missing weights. The work focused on turning pre-trained vision models into a reliable production inference service rather than manual annotation. • Implemented POST /api/ai/detect for multiple detection categories (forest fire hotspots, road cracks, humans/vehicles, photovoltaic defects). • Integrated Ultralytics/YOLO real inference and improved reliability via clear error handling when weights are missing. • Added GET /api/ai/model-status to keep the front end synchronized with model readiness. • Configured and validated model-weight paths for task-specific detection workflows.

2024 - 2024

Education

Work History