PopularMajor Project
Artificial Intelligence
Face Recognition Attendance System
Camera-based attendance with liveness detection and analytics
- Python
- OpenCV
- FaceNet
- Flask
- +2
₹4,900₹5,999
4.8/301 reviews
Artificial Intelligence · Capstone
Vehicle Detection with Swin Transformer
A transformer detector benchmarked against a CNN baseline
4-5 weeks build
Overview
A research-style comparison: a Faster R-CNN detector with a Swin Transformer backbone and the same detector with a ResNet backbone are trained on the same vehicle dataset, such as UA-DETRAC or a COCO vehicle subset. The report compares mean average precision, small-object performance and inference speed, and attention maps show where the transformer looks.
What makes it stand up
- 01Swin Transformer backbone explained window by window
- 02Identical training recipe for a fair CNN comparison
- 03mAP, small-object AP and FPS measured on both models
- 04Attention visualisations on difficult scenes
- 05Reproducible MMDetection configs for every run
Module breakdown
- 01
Data
Vehicle dataset converted to COCO format and split.
- 02
Baseline
Faster R-CNN with a ResNet backbone.
- 03
Transformer model
Same detector with a Swin backbone.
- 04
Benchmark
Accuracy and speed scripts with plots.
- 05
Explainability
Attention and error analysis on hard cases.
After this, you will be able to
- Explain shifted-window attention in plain terms
- Defend a fair experimental comparison in a viva
- Demo both detectors side by side on traffic video
Technology stack
- Python
- PyTorch
- MMDetection
- OpenCV
- Matplotlib
- Level
- Capstone
- Domain
- Artificial Intelligence
- Build time
- 4-5 weeks
- Difficulty
You receive
- Complete, commented source code
- 60-90 page project report (IEEE format)
- Editable presentation deck
- Architecture, ER and UML diagrams
- Local setup and deployment guide
- Sample dataset or seed data
- Viva question bank with answers
- One 45-minute walkthrough call
- 30 days of doubt-clearing support