# ezharjan/autoscenegen architecture

> Architecture diagram of the GitHub repository ezharjan/autoscenegen, generated by GitDiagram from its file tree, README and a sample of source files (last updated 2026-10-03). It is an AI-made overview: check details against the source.

- Interactive diagram: https://gitdiagram.com/ezharjan/autoscenegen
- Repository: https://github.com/ezharjan/autoscenegen

## Overview

AutoSceneGen is a research toolkit for turning traffic descriptions (and, in a separate image workflow, images) into simulator scenarios, collecting synthetic trajectories, and preparing datasets for motion-planning or prediction experiments. The main documented path is description → LLM-generated scenario script → CARLA execution → trajectory collection → dataset use. The excerpts show the prompt, generation, validation, and saving connections; the README documents broader simulator and dataset capabilities. Collector internals and several conversion scripts were not sampled, so their links are limited to documented relationships or omitted.

Main entry points: [`Codes/ICL/gen_scenario_by_llm.py`](https://github.com/ezharjan/autoscenegen/blob/master/Codes/ICL/gen_scenario_by_llm.py) (Single generation), [`Codes/ICL/batch_gen_scenario_by_llm.py`](https://github.com/ezharjan/autoscenegen/blob/master/Codes/ICL/batch_gen_scenario_by_llm.py) (Batch generation), [`Codes/ICL/save_content.py`](https://github.com/ezharjan/autoscenegen/blob/master/Codes/ICL/save_content.py) (Validate and save), [`Codes/ICL/img2txt.py`](https://github.com/ezharjan/autoscenegen/blob/master/Codes/ICL/img2txt.py) (Image description)

## Diagram (Mermaid)

```mermaid
flowchart TD

subgraph group_scenario_authoring["Scenario Authoring"]
  node_text_prompts["Prompt examples<br/>[questions.py]"]
  node_single_gen["Single generation"]
  node_batch_gen["Batch generation"]
  node_validate_save["Validate and save<br/>[save_content.py]"]
  node_validator["Script validation"]
end

subgraph group_simulation["CARLA Simulation"]
  node_scenario_files["Scenario scripts<br/>[0315203222.py]"]
  node_manual_control["Manual control<br/>[manual_control.py]"]
end

subgraph group_collection["Data Collection"]
  node_apollo_collector["Apollo collection"]
  node_pishgu_collector["Pishgu collection"]
  node_auto_collector["Automated collection"]
  node_collected_data["Trajectory data"]
end

subgraph group_dataset["Dataset Preparation"]
  node_log_conversion["Log conversion<br/>[log2csv.py]"]
  node_csv_data_conversion["CSV data conversion<br/>[csv2data.py]"]
  node_data_csv_conversion["Data to CSV<br/>[data2csv.py]"]
  node_msrl_conversion["MSRL conversion<br/>[pishgu2msrl.py]"]
  node_pedestrian_conversion["Pedestrian conversion<br/>[csvs2dataset.py]"]
end

subgraph group_image_inputs["Image Inputs"]
  node_image_prompt["Image prompts<br/>[questions.py]"]
  node_image_single["Image description<br/>[img2txt.py]"]
  node_image_batch["Batch image extraction"]
end

node_user(("Scenario author"))
node_openai{{"OpenAI GPT-4"}}
node_carla{{"CARLA simulator"}}
node_motion_planner(("Planner training"))
node_image_user(("Image analyst"))
node_gemini{{"Gemini Vision"}}

node_user -->|"describes scenario"| node_single_gen
node_single_gen -->|"uses prompts"| node_text_prompts
node_single_gen -->|"requests generation"| node_openai
node_single_gen -->|"saves response"| node_validate_save
node_user -->|"supplies scenarios"| node_batch_gen
node_batch_gen -->|"uses prompts"| node_text_prompts
node_batch_gen -->|"requests generation"| node_openai
node_batch_gen -->|"saves responses"| node_validate_save
node_validate_save -->|"validates code"| node_validator
node_validate_save -->|"writes scripts"| node_scenario_files
node_scenario_files -->|"executes scenarios"| node_carla
node_manual_control -->|"controls simulation"| node_carla
node_carla -.->|"provides simulated data"| node_apollo_collector
node_carla -.->|"provides simulated data"| node_pishgu_collector
node_carla -.->|"provides simulated data"| node_auto_collector
node_apollo_collector -->|"collects trajectories"| node_collected_data
node_pishgu_collector -->|"collects trajectories"| node_collected_data
node_auto_collector -.->|"collects data"| node_collected_data
node_collected_data -->|"trains or evaluates"| node_motion_planner
node_collected_data -.->|"converts logs"| node_log_conversion
node_collected_data -.->|"converts data"| node_csv_data_conversion
node_collected_data -.->|"exports CSV"| node_data_csv_conversion
node_collected_data -.->|"prepares MSRL data"| node_msrl_conversion
node_collected_data -.->|"prepares pedestrian data"| node_pedestrian_conversion
node_image_user -->|"provides image"| node_image_single
node_image_single -->|"selects prompt"| node_image_prompt
node_image_single -->|"requests description"| node_gemini
node_image_user -->|"provides images"| node_image_batch
node_image_batch -->|"selects prompt"| node_image_prompt
node_image_batch -->|"requests extraction"| node_gemini

click node_text_prompts "https://github.com/ezharjan/autoscenegen/blob/master/Codes/ICL/questions.py"
click node_single_gen "https://github.com/ezharjan/autoscenegen/blob/master/Codes/ICL/gen_scenario_by_llm.py"
click node_batch_gen "https://github.com/ezharjan/autoscenegen/blob/master/Codes/ICL/batch_gen_scenario_by_llm.py"
click node_validate_save "https://github.com/ezharjan/autoscenegen/blob/master/Codes/ICL/save_content.py"
click node_validator "https://github.com/ezharjan/autoscenegen/blob/master/Codes/ICL/validate_content.py"
click node_scenario_files "https://github.com/ezharjan/autoscenegen/blob/master/Codes/CarlaPy/scenarios/0315203222.py"
click node_manual_control "https://github.com/ezharjan/autoscenegen/blob/master/Codes/CarlaPy/manual_control.py"
click node_apollo_collector "https://github.com/ezharjan/autoscenegen/blob/master/CarlaDatasetCollector/collect_apollo_data.py"
click node_pishgu_collector "https://github.com/ezharjan/autoscenegen/blob/master/CarlaDatasetCollector/collect_pishgu_data.py"
click node_auto_collector "https://github.com/ezharjan/autoscenegen/blob/master/CarlaDatasetCollector/auto_collect_data.py"
click node_log_conversion "https://github.com/ezharjan/autoscenegen/blob/master/CarlaDatasetCollector/log2csv.py"
click node_csv_data_conversion "https://github.com/ezharjan/autoscenegen/blob/master/CarlaDatasetCollector/csv2data.py"
click node_data_csv_conversion "https://github.com/ezharjan/autoscenegen/blob/master/CarlaDatasetCollector/data2csv.py"
click node_msrl_conversion "https://github.com/ezharjan/autoscenegen/blob/master/Dataset/MSRL-Data%26Toolkits/pishgu2msrl.py"
click node_pedestrian_conversion "https://github.com/ezharjan/autoscenegen/blob/master/Dataset/Pishgu-All-Pedestrian(NGSIM-style)/csvs2dataset.py"
click node_image_prompt "https://github.com/ezharjan/autoscenegen/blob/master/Codes/ICL/questions.py"
click node_image_single "https://github.com/ezharjan/autoscenegen/blob/master/Codes/ICL/img2txt.py"
click node_image_batch "https://github.com/ezharjan/autoscenegen/blob/master/Codes/ICL/extract_info_from_imgs.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_text_prompts,node_single_gen,node_batch_gen,node_validate_save,node_validator toneBlue
class node_scenario_files,node_manual_control toneAmber
class node_apollo_collector,node_pishgu_collector,node_auto_collector,node_collected_data toneMint
class node_log_conversion,node_csv_data_conversion,node_data_csv_conversion,node_msrl_conversion,node_pedestrian_conversion toneRose
class node_image_prompt,node_image_single,node_image_batch,node_user,node_openai,node_carla,node_motion_planner,node_image_user,node_gemini toneIndigo
```

## Components

### Scenario Authoring

- [Prompt examples](https://github.com/ezharjan/autoscenegen/blob/master/Codes/ICL/questions.py): `Codes/ICL/questions.py`
- [Single generation](https://github.com/ezharjan/autoscenegen/blob/master/Codes/ICL/gen_scenario_by_llm.py): `Codes/ICL/gen_scenario_by_llm.py`
- [Batch generation](https://github.com/ezharjan/autoscenegen/blob/master/Codes/ICL/batch_gen_scenario_by_llm.py): `Codes/ICL/batch_gen_scenario_by_llm.py`
- [Validate and save](https://github.com/ezharjan/autoscenegen/blob/master/Codes/ICL/save_content.py): `Codes/ICL/save_content.py`
- [Script validation](https://github.com/ezharjan/autoscenegen/blob/master/Codes/ICL/validate_content.py): `Codes/ICL/validate_content.py`

### CARLA Simulation

- [Scenario scripts](https://github.com/ezharjan/autoscenegen/blob/master/Codes/CarlaPy/scenarios/0315203222.py): `Codes/CarlaPy/scenarios/0315203222.py`
- [Manual control](https://github.com/ezharjan/autoscenegen/blob/master/Codes/CarlaPy/manual_control.py): `Codes/CarlaPy/manual_control.py`

### Data Collection

- [Apollo collection](https://github.com/ezharjan/autoscenegen/blob/master/CarlaDatasetCollector/collect_apollo_data.py): `CarlaDatasetCollector/collect_apollo_data.py`
- [Pishgu collection](https://github.com/ezharjan/autoscenegen/blob/master/CarlaDatasetCollector/collect_pishgu_data.py): `CarlaDatasetCollector/collect_pishgu_data.py`
- [Automated collection](https://github.com/ezharjan/autoscenegen/blob/master/CarlaDatasetCollector/auto_collect_data.py): `CarlaDatasetCollector/auto_collect_data.py`
- **Trajectory data**

### Dataset Preparation

- [Log conversion](https://github.com/ezharjan/autoscenegen/blob/master/CarlaDatasetCollector/log2csv.py): `CarlaDatasetCollector/log2csv.py`
- [CSV data conversion](https://github.com/ezharjan/autoscenegen/blob/master/CarlaDatasetCollector/csv2data.py): `CarlaDatasetCollector/csv2data.py`
- [Data to CSV](https://github.com/ezharjan/autoscenegen/blob/master/CarlaDatasetCollector/data2csv.py): `CarlaDatasetCollector/data2csv.py`
- [MSRL conversion](https://github.com/ezharjan/autoscenegen/blob/master/Dataset/MSRL-Data%26Toolkits/pishgu2msrl.py): `Dataset/MSRL-Data&Toolkits/pishgu2msrl.py`
- [Pedestrian conversion](https://github.com/ezharjan/autoscenegen/blob/master/Dataset/Pishgu-All-Pedestrian(NGSIM-style)/csvs2dataset.py): `Dataset/Pishgu-All-Pedestrian(NGSIM-style)/csvs2dataset.py`

### Image Inputs

- [Image prompts](https://github.com/ezharjan/autoscenegen/blob/master/Codes/ICL/questions.py): `Codes/ICL/questions.py`
- [Image description](https://github.com/ezharjan/autoscenegen/blob/master/Codes/ICL/img2txt.py): `Codes/ICL/img2txt.py`
- [Batch image extraction](https://github.com/ezharjan/autoscenegen/blob/master/Codes/ICL/extract_info_from_imgs.py): `Codes/ICL/extract_info_from_imgs.py`

### Other components

- **Scenario author**
- **OpenAI GPT-4**
- **CARLA simulator**
- **Planner training**
- **Image analyst**
- **Gemini Vision**

## Connections

- Scenario author → Single generation: describes scenario (evidence: `README.md`)
- Single generation → Prompt examples: uses prompts (evidence: `Codes/ICL/gen_scenario_by_llm.py`)
- Single generation → OpenAI GPT-4: requests generation (evidence: `Codes/ICL/gen_scenario_by_llm.py`)
- Single generation → Validate and save: saves response (evidence: `Codes/ICL/gen_scenario_by_llm.py`)
- Scenario author → Batch generation: supplies scenarios (evidence: `README.md`)
- Batch generation → Prompt examples: uses prompts (evidence: `Codes/ICL/batch_gen_scenario_by_llm.py`)
- Batch generation → OpenAI GPT-4: requests generation (evidence: `Codes/ICL/batch_gen_scenario_by_llm.py`)
- Batch generation → Validate and save: saves responses (evidence: `Codes/ICL/batch_gen_scenario_by_llm.py`)
- Validate and save → Script validation: validates code (evidence: `Codes/ICL/save_content.py`)
- Validate and save → Scenario scripts: writes scripts (evidence: `Codes/ICL/save_content.py`)
- Scenario scripts → CARLA simulator: executes scenarios (evidence: `README.md`)
- Manual control → CARLA simulator: controls simulation (evidence: `Codes/CarlaPy/manual_control.py`)
- CARLA simulator → Apollo collection: provides simulated data
- CARLA simulator → Pishgu collection: provides simulated data
- CARLA simulator → Automated collection: provides simulated data
- Apollo collection → Trajectory data: collects trajectories (evidence: `README.md`)
- Pishgu collection → Trajectory data: collects trajectories (evidence: `README.md`)
- Automated collection → Trajectory data: collects data
- Trajectory data → Planner training: trains or evaluates (evidence: `README.md`)
- Trajectory data → Log conversion: converts logs
- Trajectory data → CSV data conversion: converts data
- Trajectory data → Data to CSV: exports CSV
- Trajectory data → MSRL conversion: prepares MSRL data
- Trajectory data → Pedestrian conversion: prepares pedestrian data
- Image analyst → Image description: provides image (evidence: `Codes/ICL/img2txt.py`)
- Image description → Image prompts: selects prompt (evidence: `Codes/ICL/img2txt.py`)
- Image description → Gemini Vision: requests description (evidence: `Codes/ICL/img2txt.py`)
- Image analyst → Batch image extraction: provides images (evidence: `Codes/ICL/extract_info_from_imgs.py`)
- Batch image extraction → Image prompts: selects prompt (evidence: `Codes/ICL/extract_info_from_imgs.py`)
- Batch image extraction → Gemini Vision: requests extraction (evidence: `Codes/ICL/extract_info_from_imgs.py`)
