ArlingtonCL2/Barkopedia_Individual_Dog_Recognition_Dataset

This dataset is for Barkopedia Challenge https://uta-acl2.github.io/barkopedia.html

📦 Dataset Description

Check Training Data here: ArlingtonCL2/Barkopedia_Individual_Dog_Recognition_Dataset

This challenge provides a dataset of labeled dog bark audio clips: 8924

• Training set: 7137 clips (~120 clips per 60 dog IDs).
• Test set: 1787 clips (~30 clips per 60 dog IDs) with hidden labels:
  • 709 public (~40%) for live leaderboard updates.
  • 1078 private (~60%) for final evaluation. You will have access to all 1787 clips in the Dataset, the result of the 709 public clips will show the evaluation result.

🔖 Labels

Each clip is annotated with: a dog id from index number 1-60.

The labels were manually generated and verified by us.

---

The train_labels.csv file provides the ground truth (correct labels) for the training set. It contains the following columns:

• audio_id: The filename of the dog bark audio clip (e.g., "bark_752352").
• pred_dog_id: The ID of the dog that the bark belongs to (an integer between 1 and 60).

For example, a row with audio_id as "bark_2408" and pred_dog_id as "17" indicates that the audio clip "bark_2408.wav" contains a bark from dog number 17.

---