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
Underwater Image Enhancement with GANs
Restores colour and contrast lost to water in underwater photographs
5-6 weeks build
Overview
Water absorbs red light first, leaving photos blue-green and flat. This kit trains a conditional GAN with a U-Net generator and PatchGAN discriminator on the paired EUVP dataset, and benchmarks it against classical fixes — white balance, CLAHE and fusion-based enhancement — on the UIEB set using PSNR, SSIM and the no-reference UIQM score.
What makes it stand up
- 01Paired image-to-image GAN with a perceptual loss term
- 02Classical baselines implemented for an honest comparison
- 03Full-reference and no-reference quality metrics side by side
- 04Before-and-after slider in the demo app
- 05Ablation of each loss term documented in the report
Module breakdown
- 01
Classical enhancement
Grey-world white balance, CLAHE and multi-scale fusion.
- 02
Generator
U-Net generator with residual blocks in the bottleneck.
- 03
Discriminator
PatchGAN scoring realism over local patches.
- 04
Training
Adversarial, L1 and VGG perceptual losses with checkpoints.
- 05
Evaluation app
Streamlit slider plus PSNR, SSIM and UIQM tables.
After this, you will be able to
- Explain why red channels fade underwater and how the GAN compensates
- Defend the choice of both reference and no-reference metrics
- Demo enhancement on dive photos the examiner has not seen
Technology stack
- Python
- PyTorch
- OpenCV
- scikit-image
- Streamlit
- Level
- Capstone
- Domain
- Artificial Intelligence
- Build time
- 5-6 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