How akshaydudhane16/bipnet works
This repository implements BIPNet burst-image restoration, with separate denoising and super-resolution code paths. The user-facing workflow is to provide a burst and task-appropriate metadata or dataset; the model restores an image, which can then be postprocessed and evaluated. Super-resolution also supports real BurstSR data and synthetic-burst generation; the README describes denoising, super-resolution, and low-light enhancement as target tasks. Some subsystem links are documented only by the source index, so unsampled internal call paths are omitted rather than inferred.
Main entry points: BIPNet_Track_2_training.py Track Training · Color_denoising_testing.py Color Testing · Grayscale_denoising_testing.py Grayscale Testing · postprocessing_functions.py Denoising Postprocess
Components 25
Burst Denoising
- Color Testing
Burst De-noising/Color_denoising_testing.py - Grayscale Testing
Burst De-noising/Grayscale_denoising_testing.py - Denoising BIPNet
Burst De-noising/Network.py - Color Test Data
Burst De-noising/dataset/color_denoise_test_set.py - Grayscale Test Data
Burst De-noising/dataset/grayscale_denoise_test_set.py - Denoising Camera Pipeline
Burst De-noising/data/camera_pipeline.py - Denoising Postprocess
Burst De-noising/data/postprocessing_functions.py - Denoising Metrics
Burst De-noising/utils/metrics.py
Burst Super-Resolution
- Track Training
Burst_Super_Resolution/BIPNet_Track_2_training.py - Track 1 Training
Burst_Super_Resolution/BIPNet_Track_1_training.py - Track Evaluation
Burst_Super_Resolution/Track_2_evaluation.py - Track 1 Evaluation
Burst_Super_Resolution/Track_1_evaluation.py - Super-Resolution BIPNet
Burst_Super_Resolution/Network.py
Data Preparation
- Real BurstSR Data
Burst_Super_Resolution/datasets/burstsr_dataset.py - Synthetic Train Data
Burst_Super_Resolution/datasets/synthetic_burst_train_set.py - Synthetic Validation Data
Burst_Super_Resolution/datasets/synthetic_burst_val_set.py - Zurich RAW Data
Burst_Super_Resolution/datasets/zurich_raw2rgb_dataset.py - Burst Synthesis
Burst_Super_Resolution/data_processing/synthetic_burst_generation.py - SR Camera Pipeline
Burst_Super_Resolution/data_processing/camera_pipeline.py
Restoration Utilities
- SR Postprocess
Burst_Super_Resolution/utils/postprocessing_functions.py - SR Metrics And Alignment
Burst_Super_Resolution/utils/metrics.py - Spatial Color Alignment
Burst_Super_Resolution/utils/spatial_color_alignment.py - Image Warping
Burst_Super_Resolution/utils/warp.py - Image Format Utilities
Burst De-noising/utils/data_format_utils.py
Other components
- Researcher
Connections 33
- Researcher → Color Testing: runs
README.md - Researcher → Grayscale Testing: runs
README.md - Color Testing → Color Test Data: loads
Burst De-noising/Color_denoising_testing.py - Color Testing → Denoising BIPNet: runs
Burst De-noising/Color_denoising_testing.py - Color Testing → Denoising Postprocess: postprocesses
Burst De-noising/Color_denoising_testing.py - Color Testing → Denoising Metrics: evaluates
Burst De-noising/Color_denoising_testing.py - Grayscale Testing → Grayscale Test Data: loads
Burst De-noising/Grayscale_denoising_testing.py - Grayscale Testing → Denoising BIPNet: runs
Burst De-noising/Grayscale_denoising_testing.py - Grayscale Testing → Denoising Postprocess: postprocesses
Burst De-noising/Grayscale_denoising_testing.py - Grayscale Testing → Denoising Metrics: evaluates
Burst De-noising/Grayscale_denoising_testing.py - Researcher → Track Training: trains
README.md - Researcher → Track Evaluation: evaluates
README.md - Track Training → Super-Resolution BIPNet: trains
Burst_Super_Resolution/BIPNet_Track_2_training.py - Track Training → Real BurstSR Data: loads
Burst_Super_Resolution/BIPNet_Track_2_training.py - Track Training → SR Postprocess: uses
Burst_Super_Resolution/BIPNet_Track_2_training.py - Track Training → SR Metrics And Alignment: measures
Burst_Super_Resolution/BIPNet_Track_2_training.py - Track Training → Image Warping: uses
Burst_Super_Resolution/BIPNet_Track_2_training.py - Track Training → SR Camera Pipeline: uses
Burst_Super_Resolution/BIPNet_Track_2_training.py - Track 1 Training → Super-Resolution BIPNet: trains
Burst_Super_Resolution/BIPNet_Track_1_training.py - Track 1 Training → Synthetic Train Data: loads
Burst_Super_Resolution/BIPNet_Track_1_training.py - Track 1 Training → Zurich RAW Data: loads
Burst_Super_Resolution/BIPNet_Track_1_training.py - Track Evaluation → Real BurstSR Data: loads
Burst_Super_Resolution/Track_2_evaluation.py - Track Evaluation → SR Postprocess: postprocesses
Burst_Super_Resolution/Track_2_evaluation.py - Track 1 Evaluation → Synthetic Validation Data: loads
Burst_Super_Resolution/Track_1_evaluation.py - Track 1 Evaluation → Super-Resolution BIPNet: evaluates
Burst_Super_Resolution/Track_1_evaluation.py - Synthetic Train Data → Burst Synthesis: generates
Burst_Super_Resolution/datasets/synthetic_burst_train_set.py - Burst Synthesis → SR Camera Pipeline: processes
Burst_Super_Resolution/data_processing/synthetic_burst_generation.py - SR Postprocess → SR Camera Pipeline: uses
Burst_Super_Resolution/utils/postprocessing_functions.py - SR Postprocess → Image Format Utilities: converts
Burst_Super_Resolution/utils/postprocessing_functions.py - SR Metrics And Alignment → Spatial Color Alignment: aligns
Burst_Super_Resolution/utils/metrics.py - SR Metrics And Alignment → Image Warping: warps
Burst_Super_Resolution/utils/metrics.py - Denoising Postprocess → Denoising Camera Pipeline: uses
Burst De-noising/data/postprocessing_functions.py - Denoising Postprocess → Image Format Utilities: converts
Burst De-noising/data/postprocessing_functions.py