Quickstart

Get started with Avala by creating your first annotation project.

Prerequisites

• An Avala account (sign up )
• Images or data to annotate (or use sample data)

Step 1: Create a Dataset

1. Log in to Mission Control 
2. Click Datasets in the left sidebar
3. Click Add Dataset
4. Enter a name and description
5. Choose visibility (public or private)

Step 2: Upload Data

After creating your dataset:

1. Click Add Items or drag files into the upload area
2. Supported formats:
   • Images: JPEG, PNG, WebP
   • Videos: MP4, MOV (converted to frame sequences)
   • Point clouds: PCD, PLY, or via MCAP
3. Wait for processing to complete

Step 3: Create a Project

Projects define your annotation workflow:

1. Navigate to Projects → Add Project
2. Select your dataset
3. Choose annotation type:
   • Bounding Box: 2D object detection
   • Polygon: Instance segmentation
   • Cuboid: 3D object detection (for point clouds)
   • Segmentation: Semantic segmentation
4. Define your label taxonomy (object classes)
5. Configure quality control settings

Step 4: Annotate Data

Open an item to start annotating:

1. Go to your dataset → Items
2. Click on an item to open the annotation editor
3. Use the toolbar to select annotation tools:
   • Box tool: Click and drag to draw bounding boxes
   • Polygon tool: Click to add vertices, close the shape
   • Select tool: Click annotations to edit or delete
4. Assign labels from your taxonomy
5. Click Save to submit your annotation

Keyboard Shortcuts

Step 5: Review Annotations

Quality control ensures annotation accuracy:

1. Navigate to Projects → your project → Review
2. Review pending annotations
3. Accept correct annotations
4. Reject and provide feedback for corrections
5. Track quality metrics in the dashboard

Step 6: Export Results

Export your labeled data:

1. Go to Exports → Create Export
2. Select your project or dataset
3. Choose export format:
   • COCO JSON: Standard object detection format
   • KITTI: Autonomous driving format
   • Avala Format: Native format with full metadata
4. Download when ready

Using the API

You can also work programmatically:

import requests
 
API_KEY = "your-api-key"
BASE_URL = "https://api.avala.ai/api/v1"
 
headers = {"X-Avala-Api-Key": API_KEY}
 
# List your datasets
response = requests.get(f"{BASE_URL}/datasets/", headers=headers)
datasets = response.json()
 
for dataset in datasets["results"]:
    print(f"{dataset['uid']}: {dataset['name']}")

# List your datasets
curl https://api.avala.ai/api/v1/datasets/ \
  -H "X-Avala-Api-Key: your-api-key"

Next Steps

• Learn about Core Concepts in depth
• Explore Mission Control features
• Read the API Reference for programmatic access
• Set up Organizations for team collaboration