# goutamgmb/deep-burst-sr architecture

> Architecture diagram of the GitHub repository goutamgmb/deep-burst-sr, generated by GitDiagram from its file tree, README and a sample of source files (last updated 2026-10-01). It is an AI-made overview: check details against the source.

- Interactive diagram: https://gitdiagram.com/goutamgmb/deep-burst-sr
- Repository: https://github.com/goutamgmb/deep-burst-sr

## Overview

Deep-Burst-SR is a RAW burst super-resolution research toolkit. Its main workflow prepares real or synthetic burst data, trains a model that aligns and fuses frames into a denoised high-resolution RGB image, and evaluates or saves predictions. The graph also includes documented dataset and evaluation branches. Some subsystem connections are documented in the README or SOURCE INDEX rather than visible excerpts; those relationships are cited accordingly.

Main entry points: [`train_settings/dbsr/default_synthetic.py`](https://github.com/goutamgmb/deep-burst-sr/blob/master/train_settings/dbsr/default_synthetic.py) (Training settings), [`evaluation/burstsr/compute_score.py`](https://github.com/goutamgmb/deep-burst-sr/blob/master/evaluation/burstsr/compute_score.py) (BurstSR evaluation), [`evaluation/synburst/compute_score.py`](https://github.com/goutamgmb/deep-burst-sr/blob/master/evaluation/synburst/compute_score.py) (Synthetic evaluation), [`models/dbsr/dbsrnet.py`](https://github.com/goutamgmb/deep-burst-sr/blob/master/models/dbsr/dbsrnet.py) (DBSR network)

## Diagram (Mermaid)

```mermaid
flowchart TD

subgraph group_data["Data preparation"]
  node_synthetic_data["Synthetic processing<br/>[processing.py]"]
  node_burst_generation["Burst generation"]
  node_camera_pipeline["Camera simulation<br/>[camera_pipeline.py]"]
  node_transforms["Data transforms<br/>[transforms.py]"]
  node_zurich["Zurich RAW2RGB<br/>[zurichraw2rgb.py]"]
  node_burstsr["BurstSR dataset<br/>[burstsr_dataset.py]"]
  node_synburst_val["Synthetic validation"]
end

subgraph group_model["Super-resolution model"]
  node_dbsr["DBSR network<br/>[dbsrnet.py]"]
  node_alignment["Optical-flow alignment<br/>[pwcnet.py]"]
  node_correlation["Correlation layer<br/>[correlation.py]"]
  node_merging["Frame fusion<br/>[merging.py]"]
  node_losses["Training losses"]
end

subgraph group_training["Training"]
  node_train_entry["Training entry<br/>[run_training.py]"]
  node_train_settings["Training settings"]
  node_trainer["Trainer<br/>[base_trainer.py]"]
  node_actor["DBSR actors<br/>[dbsr_actors.py]"]
end

subgraph group_evaluation["Evaluation"]
  node_network_params["Evaluation networks<br/>[network_param.py]"]
  node_burst_eval["BurstSR evaluation<br/>[compute_score.py]"]
  node_syn_eval["Synthetic evaluation<br/>[compute_score.py]"]
  node_save_results["Save predictions"]
  node_visualize["Visualize results"]
  node_metrics["Quality metrics"]
end

node_researcher(("Researcher"))
node_environment["Environment settings<br/>[environment.py]"]

node_researcher -->|"starts training"| node_train_entry
node_train_entry -->|"loads settings"| node_train_settings
node_train_settings -->|"selects dataset"| node_zurich
node_train_settings -->|"configures processing"| node_synthetic_data
node_train_settings -->|"configures transforms"| node_transforms
node_zurich -->|"provides images"| node_synthetic_data
node_synthetic_data -->|"generates bursts"| node_burst_generation
node_burst_generation -->|"simulates camera"| node_camera_pipeline
node_train_settings -->|"configures training"| node_trainer
node_trainer -->|"runs training actor"| node_actor
node_actor -->|"runs network"| node_dbsr
node_actor -->|"computes loss"| node_losses
node_dbsr -->|"aligns frames"| node_alignment
node_alignment -->|"uses correlation"| node_correlation
node_dbsr -->|"fuses frames"| node_merging
node_burstsr -->|"gets data path"| node_environment
node_synburst_val -->|"gets data path"| node_environment
node_researcher -->|"evaluates BurstSR"| node_burst_eval
node_researcher -->|"evaluates synthetic set"| node_syn_eval
node_burst_eval -->|"reads dataset"| node_burstsr
node_syn_eval -->|"reads dataset"| node_synburst_val
node_burst_eval -->|"aligns predictions"| node_alignment
node_burst_eval -->|"computes scores"| node_metrics
node_syn_eval -->|"computes scores"| node_metrics
node_burst_eval -.->|"loads experiment networks"| node_network_params
node_syn_eval -.->|"loads experiment networks"| node_network_params
node_researcher -->|"saves predictions"| node_save_results
node_researcher -->|"views results"| node_visualize

click node_train_entry "https://github.com/goutamgmb/deep-burst-sr/blob/master/run_training.py"
click node_train_settings "https://github.com/goutamgmb/deep-burst-sr/blob/master/train_settings/dbsr/default_synthetic.py"
click node_trainer "https://github.com/goutamgmb/deep-burst-sr/blob/master/trainers/base_trainer.py"
click node_actor "https://github.com/goutamgmb/deep-burst-sr/blob/master/actors/dbsr_actors.py"
click node_synthetic_data "https://github.com/goutamgmb/deep-burst-sr/blob/master/data/processing.py"
click node_burst_generation "https://github.com/goutamgmb/deep-burst-sr/blob/master/data/synthetic_burst_generation.py"
click node_camera_pipeline "https://github.com/goutamgmb/deep-burst-sr/blob/master/data/camera_pipeline.py"
click node_transforms "https://github.com/goutamgmb/deep-burst-sr/blob/master/data/transforms.py"
click node_zurich "https://github.com/goutamgmb/deep-burst-sr/blob/master/dataset/zurichraw2rgb.py"
click node_burstsr "https://github.com/goutamgmb/deep-burst-sr/blob/master/dataset/burstsr_dataset.py"
click node_synburst_val "https://github.com/goutamgmb/deep-burst-sr/blob/master/dataset/synthetic_burst_val_set.py"
click node_dbsr "https://github.com/goutamgmb/deep-burst-sr/blob/master/models/dbsr/dbsrnet.py"
click node_alignment "https://github.com/goutamgmb/deep-burst-sr/blob/master/models/alignment/pwcnet.py"
click node_correlation "https://github.com/goutamgmb/deep-burst-sr/blob/master/external/pwcnet/correlation/correlation.py"
click node_merging "https://github.com/goutamgmb/deep-burst-sr/blob/master/models/dbsr/merging.py"
click node_losses "https://github.com/goutamgmb/deep-burst-sr/blob/master/models/loss/image_quality_v2.py"
click node_network_params "https://github.com/goutamgmb/deep-burst-sr/blob/master/evaluation/common_utils/network_param.py"
click node_burst_eval "https://github.com/goutamgmb/deep-burst-sr/blob/master/evaluation/burstsr/compute_score.py"
click node_syn_eval "https://github.com/goutamgmb/deep-burst-sr/blob/master/evaluation/synburst/compute_score.py"
click node_save_results "https://github.com/goutamgmb/deep-burst-sr/tree/master/evaluation"
click node_visualize "https://github.com/goutamgmb/deep-burst-sr/tree/master/evaluation"
click node_metrics "https://github.com/goutamgmb/deep-burst-sr/blob/master/models/loss/image_quality_v2.py"
click node_environment "https://github.com/goutamgmb/deep-burst-sr/blob/master/admin/environment.py"

classDef toneNeutral fill:#f8fafc,stroke:#334155,stroke-width:1.5px,color:#0f172a
classDef toneBlue fill:#dbeafe,stroke:#2563eb,stroke-width:1.5px,color:#172554
classDef toneAmber fill:#fef3c7,stroke:#d97706,stroke-width:1.5px,color:#78350f
classDef toneMint fill:#dcfce7,stroke:#16a34a,stroke-width:1.5px,color:#14532d
classDef toneRose fill:#ffe4e6,stroke:#e11d48,stroke-width:1.5px,color:#881337
classDef toneIndigo fill:#e0e7ff,stroke:#4f46e5,stroke-width:1.5px,color:#312e81
classDef toneTeal fill:#ccfbf1,stroke:#0f766e,stroke-width:1.5px,color:#134e4a
class node_synthetic_data,node_burst_generation,node_camera_pipeline,node_transforms,node_zurich,node_burstsr,node_synburst_val toneBlue
class node_dbsr,node_alignment,node_correlation,node_merging,node_losses toneAmber
class node_train_entry,node_train_settings,node_trainer,node_actor toneMint
class node_network_params,node_burst_eval,node_syn_eval,node_save_results,node_visualize,node_metrics toneRose
class node_researcher toneIndigo
class node_environment toneTeal
```

## Components

### Data preparation

- [Synthetic processing](https://github.com/goutamgmb/deep-burst-sr/blob/master/data/processing.py): `data/processing.py`
- [Burst generation](https://github.com/goutamgmb/deep-burst-sr/blob/master/data/synthetic_burst_generation.py): `data/synthetic_burst_generation.py`
- [Camera simulation](https://github.com/goutamgmb/deep-burst-sr/blob/master/data/camera_pipeline.py): `data/camera_pipeline.py`
- [Data transforms](https://github.com/goutamgmb/deep-burst-sr/blob/master/data/transforms.py): `data/transforms.py`
- [Zurich RAW2RGB](https://github.com/goutamgmb/deep-burst-sr/blob/master/dataset/zurichraw2rgb.py): `dataset/zurichraw2rgb.py`
- [BurstSR dataset](https://github.com/goutamgmb/deep-burst-sr/blob/master/dataset/burstsr_dataset.py): `dataset/burstsr_dataset.py`
- [Synthetic validation](https://github.com/goutamgmb/deep-burst-sr/blob/master/dataset/synthetic_burst_val_set.py): `dataset/synthetic_burst_val_set.py`

### Super-resolution model

- [DBSR network](https://github.com/goutamgmb/deep-burst-sr/blob/master/models/dbsr/dbsrnet.py): `models/dbsr/dbsrnet.py`
- [Optical-flow alignment](https://github.com/goutamgmb/deep-burst-sr/blob/master/models/alignment/pwcnet.py): `models/alignment/pwcnet.py`
- [Correlation layer](https://github.com/goutamgmb/deep-burst-sr/blob/master/external/pwcnet/correlation/correlation.py): `external/pwcnet/correlation/correlation.py`
- [Frame fusion](https://github.com/goutamgmb/deep-burst-sr/blob/master/models/dbsr/merging.py): `models/dbsr/merging.py`
- [Training losses](https://github.com/goutamgmb/deep-burst-sr/blob/master/models/loss/image_quality_v2.py): `models/loss/image_quality_v2.py`

### Training

- [Training entry](https://github.com/goutamgmb/deep-burst-sr/blob/master/run_training.py): `run_training.py`
- [Training settings](https://github.com/goutamgmb/deep-burst-sr/blob/master/train_settings/dbsr/default_synthetic.py): `train_settings/dbsr/default_synthetic.py`
- [Trainer](https://github.com/goutamgmb/deep-burst-sr/blob/master/trainers/base_trainer.py): `trainers/base_trainer.py`
- [DBSR actors](https://github.com/goutamgmb/deep-burst-sr/blob/master/actors/dbsr_actors.py): `actors/dbsr_actors.py`

### Evaluation

- [Evaluation networks](https://github.com/goutamgmb/deep-burst-sr/blob/master/evaluation/common_utils/network_param.py): `evaluation/common_utils/network_param.py`
- [BurstSR evaluation](https://github.com/goutamgmb/deep-burst-sr/blob/master/evaluation/burstsr/compute_score.py): `evaluation/burstsr/compute_score.py`
- [Synthetic evaluation](https://github.com/goutamgmb/deep-burst-sr/blob/master/evaluation/synburst/compute_score.py): `evaluation/synburst/compute_score.py`
- [Save predictions](https://github.com/goutamgmb/deep-burst-sr/tree/master/evaluation): `evaluation`
- [Visualize results](https://github.com/goutamgmb/deep-burst-sr/tree/master/evaluation): `evaluation`
- [Quality metrics](https://github.com/goutamgmb/deep-burst-sr/blob/master/models/loss/image_quality_v2.py): `models/loss/image_quality_v2.py`

### Other components

- **Researcher**
- [Environment settings](https://github.com/goutamgmb/deep-burst-sr/blob/master/admin/environment.py): `admin/environment.py`

## Connections

- Researcher → Training entry: starts training
- Training entry → Training settings: loads settings (evidence: `run_training.py`)
- Training settings → Zurich RAW2RGB: selects dataset (evidence: `train_settings/dbsr/default_synthetic.py`)
- Training settings → Synthetic processing: configures processing (evidence: `train_settings/dbsr/default_synthetic.py`)
- Training settings → Data transforms: configures transforms (evidence: `train_settings/dbsr/default_synthetic.py`)
- Zurich RAW2RGB → Synthetic processing: provides images
- Synthetic processing → Burst generation: generates bursts (evidence: `data/processing.py`)
- Burst generation → Camera simulation: simulates camera (evidence: `data/synthetic_burst_generation.py`)
- Training settings → Trainer: configures training (evidence: `train_settings/dbsr/default_synthetic.py`)
- Trainer → DBSR actors: runs training actor
- DBSR actors → DBSR network: runs network (evidence: `actors/dbsr_actors.py`)
- DBSR actors → Training losses: computes loss (evidence: `actors/dbsr_actors.py`)
- DBSR network → Optical-flow alignment: aligns frames (evidence: `models/dbsr/dbsrnet.py`)
- Optical-flow alignment → Correlation layer: uses correlation (evidence: `models/alignment/pwcnet.py`)
- DBSR network → Frame fusion: fuses frames (evidence: `models/dbsr/dbsrnet.py`)
- BurstSR dataset → Environment settings: gets data path (evidence: `dataset/burstsr_dataset.py`)
- Synthetic validation → Environment settings: gets data path (evidence: `dataset/synthetic_burst_val_set.py`)
- Researcher → BurstSR evaluation: evaluates BurstSR
- Researcher → Synthetic evaluation: evaluates synthetic set
- BurstSR evaluation → BurstSR dataset: reads dataset (evidence: `evaluation/burstsr/compute_score.py`)
- Synthetic evaluation → Synthetic validation: reads dataset (evidence: `evaluation/synburst/compute_score.py`)
- BurstSR evaluation → Optical-flow alignment: aligns predictions (evidence: `evaluation/burstsr/compute_score.py`)
- BurstSR evaluation → Quality metrics: computes scores (evidence: `evaluation/burstsr/compute_score.py`)
- Synthetic evaluation → Quality metrics: computes scores (evidence: `evaluation/synburst/compute_score.py`)
- BurstSR evaluation → Evaluation networks: loads experiment networks
- Synthetic evaluation → Evaluation networks: loads experiment networks
- Researcher → Save predictions: saves predictions
- Researcher → Visualize results: views results
