You will get precise image recognition that meets your requirements
Project details
Highlights:
✅ Lightweight Model: Only 45.2 MB, 11.8M parameters
✅ High Accuracy: mIoU 85.4%, outperforms traditional methods
✅ Image Enhancement: Retinex algorithm, 15-20% improvement in low-light scenarios
✅ Multi-Dataset Validation: Validated on concrete, metal, wood, and other materials
✅ Fast Inference: < 2 seconds per image (GPU)
✅ Batch Processing: Automatic multi-folder processing
I’m a Computer Vision Engineer specializing in building high-performance models.
Over the last 3 years, I’ve focused heavily on pushing the boundaries of accuracy. My track record includes delivering multiple projects that achieved State-of-the-Art (SOTA) results.
I combine strong research capabilities with solid engineering skills in computer vision deployment. I’m looking for a challenging role where I can continue to drive algorithmic innovation and solve complex visual problems."
✅ Lightweight Model: Only 45.2 MB, 11.8M parameters
✅ High Accuracy: mIoU 85.4%, outperforms traditional methods
✅ Image Enhancement: Retinex algorithm, 15-20% improvement in low-light scenarios
✅ Multi-Dataset Validation: Validated on concrete, metal, wood, and other materials
✅ Fast Inference: < 2 seconds per image (GPU)
✅ Batch Processing: Automatic multi-folder processing
I’m a Computer Vision Engineer specializing in building high-performance models.
Over the last 3 years, I’ve focused heavily on pushing the boundaries of accuracy. My track record includes delivering multiple projects that achieved State-of-the-Art (SOTA) results.
I combine strong research capabilities with solid engineering skills in computer vision deployment. I’m looking for a challenging role where I can continue to drive algorithmic innovation and solve complex visual problems."
Machine Learning Tools
NumPy, Python, PyTorch, scikit-learn, Stanford CoreNLP, TensorFlowWhat's included $100
These options are included with the project scope.
$100
- Delivery Time 20 days
- Number of Revisions 0
- Number of Model Variations 1
- Number of Scenarios 0
- Number of Graphs/Charts 10
- Model Validation/Testing
- Model Documentation
- Data Source Connectivity
- Source Code
Optional add-ons
You can add these on the next page.
Additional Scenario
+$10
Additional Graph/Chart
+$10Frequently asked questions
About Zhu
AI Engineer | Computer Vision, RAG & Data Pipelines
Shanghai, China - 10:38 pm local time
What I can help with:
- Computer Vision: object detection, segmentation, medical/industrial image analysis, YOLO/SAM/ViT/Swin/PyTorch workflows
- RAG & AI chat: document ingestion, vector search, reranking, citations, answerability checks, hallucination reduction
- Data pipelines: structured extraction, ClickHouse/Qdrant storage, API services, dashboards, monitoring, quality review
- Deployment: FastAPI, Next.js, Docker, ONNX/NCNN, Android/edge AI prototypes
Proof from recent projects:
- Low-light crack segmentation: 45.2MB model, 11.8M parameters, mIoU 85.4%, F1 90.9%, Android deployment path
- Medical dermatology classification: HAM10000 Acc 98.90%, Macro F1 0.984, Grad-CAM review
- Evidence-first RAG/data intelligence: millions of papers, news, Reddit posts, comments, and job records with cited retrieval
My working style:
I define success metrics first, build a reproducible prototype, show intermediate results early, and deliver code, documentation, screenshots/demo, and a handoff plan.
If your project needs a working AI system with evaluation metrics and clean delivery, send me your dataset/source, target output, and acceptance criteria. I’ll propose a practical technical plan.
Steps for completing your project
After purchasing the project, send requirements so Zhu can start the project.
Delivery time starts when Zhu receives requirements from you.
Zhu works on your project following the steps below.
Revisions may occur after the delivery date.
Data Preparation & Preprocessing
Data collection guidance, image enhancement, annotation, augmentation
Model Development & Training
Algorithm design, model training, evaluation, optimization