Once bitten, twice shy: Aggressive and defeated pigs begin agonistic encounters with more negative emotions

Pattern Mining-Based Pig Behavior Analysis for Health and Welfare Monitoring

The increasing popularity of pigs has prompted farmers to increase pig production to meet the growing demand. However, while the number of pigs is increasing, that of farm workers has been declining, making it challenging to perform various farm tasks, the most important among them being managing the pigs' health and welfare. This study proposes a pattern mining-based pig behavior analysis system to provide visualized information and behavioral patterns, assisting farmers in effectively monitoring and assessing pigs' health and welfare. The system consists of four modules: (1) data acquisition module for collecting pigs video; (2) detection and tracking module for localizing and uniquely identifying pigs, using tracking information to crop pig images; (3) pig behavior recognition module for recognizing pig behaviors from sequences of cropped images; and (4) pig behavior analysis module for providing visualized information and behavioral patterns to effectively help farmers understand and manage pigs. In the second module, we utilize ByteTrack, which comprises YOLOx as the detector and the BYTE algorithm as the tracker, while MnasNet and LSTM serve as appearance features and temporal information extractors in the third module. The experimental results show that the system achieved a multi-object tracking accuracy of 0.971 for tracking and an F1 score of 0.931 for behavior recognition, while also highlighting the effectiveness of visualization and pattern mining in helping farmers comprehend and manage pigs' health and welfare.

A systematic approach to defining and verifying descriptors used in the Qualitative Behavioural Assessment of sows

Qualitative Behaviour Assessment (QBA) is a welfare evaluation tool that uses a holistic approach to capturing an animal's emotional state. Lists of QBA descriptors validated to assess pig welfare exist, but their definitions are often not described in peer-reviewed literature and the processes used to develop definitions are lacking. The objective of this study is to detail a systematic approach to creating clear definitions for a pre-existing fixed list of QBA descriptors and test their application. A fixed list of 20 descriptors from the EU Welfare Quality® assessment protocol for pigs was modified, and ten pig experts were recruited to assist with defining these descriptors in a focus group-style discussion. Half of the experts involved in creating descriptor definitions partook in a subsequent step, where the newly developed definitions were tested by implementing QBA on a video library of post-weaned sows selected to capture the breadth of sow behaviour. Experts displayed excellent agreement in identifying a PCA dimension interpreted as the valence of descriptors and good agreement for another reflecting arousal. Inter-observer reliability was also measured for each descriptor. Only two descriptors exhibited less than moderate agreement between experts whereas half of the descriptors evoked substantial agreement or better. These findings support our process to delineate clear definitions for a fixed list of QBA descriptors in pigs. This study is the first of its kind detailing the in-depth process of creating and verifying descriptor definitions for future use in sow welfare assessment.