DIIP-Dataset for Pig interaction identification with an Enriched Environment

Pig's Interaction with Enrichment

Interaction type                                               Description

Interacting with standard enrichment          Pig actively sniffs, noses, bites, chews, or holds a plastic toy or softwood in a mouth

Interacting with jute bags                            Pig actively sniffs, noses, bite, chews, or hold a jute bag in their mouth (observed only in enriched pigs)

Feeding                                                       Pig's head in fodder beet, eating or chewing

We annotated 6 different class labels (Feeding, Jute bag, Pig Head, Snout, Toy, Wooden Beam)with a bounding box in each image.

parameters:

            class-label – integer value representing the object class starting from 0, 1, 2, 3, 4, 5

            respective bounding box coordinates: x_center, y_center, width, height

Dataset folder Structure:

    The directory structure is as shown below:

          xml_annotated – folder containing images and labels (XML)

         yolo_annotated – folder containing images and labels (YOLO).

This annotation label directory is compatible with XML, YOLOv4, and YOLOv7 without modification. There are 7187 plus corresponding bounding boxes files.

There is also a data preprocessing file that can be used to process the data to create training metadata either for YOLOv4 or YOLOv7.

Funding:

This research is part of the EU-China HealthyLivestock project funded by the European Union H2020 research and innovation program under grant agreement number 773436. The European Commission’s support to produce this publication does not constitute an endorsement of the contents, which reflect the views only of the authors, and the Commission cannot be held responsible for any use which may be made of the information contained therein.

Acknowledgment:

We are grateful to Stephanie M. Matheson and Katarína Bučková both of QUB for help with image annotation.