Dataset Viewer
Auto-converted to Parquet Duplicate
Search is not available for this dataset
image
image
label
class label
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
null
End of preview. Expand in Data Studio

Apex Legends Enemy Detection Dataset

Dataset for detecting players in Apex Legends gameplay footage.
4 921 frames — 2 classes: enemy, mate.

Dataset Structure

Split Images Labels
train 3 937 3 937
val 984 984
images/
  train/   # 3937 × .png
  val/     # 984  × .png
labels/    # YOLO .txt, mirrors images/
dataset.yaml

Annotation Format

YOLO — each .txt contains one row per bounding box:

<class_id> <cx> <cy> <w> <h>   # normalized [0..1]
class_id label
0 enemy
1 mate

Source

Frames extracted from Apex Legends gameplay recordings (DVR captures).
Labeled manually using CVAT.

License

CC BY-NC 4.0non-commercial use only.

Apex Legends is a trademark of Electronic Arts Inc. / Respawn Entertainment.
All game assets depicted in the images remain the property of their respective owners.
This dataset is intended solely for academic research and non-commercial purposes.

Downloads last month
39