每天早上一封邮件，把昨天的 AI 梳理好订阅邮件

METAL LAB

Interaction valence reveals contrasting social networks in dairy cattle

arXiv:2608.192222026-08-21

把奶牛之间的互动分成友好和敌对两类后,牛群的社交网络图完全变了样

研究团队利用现有的姿态追踪视频分析技术,对一家商业奶牛场挤奶前等候区7小时39分钟的监控视频进行分析,记录了牛与牛之间的互动关系。他们把这些互动分为友好行为(舔舐、梳理)和敌对行为(驱赶、顶撞)两类,分别绘制社交网络后发现两者结构明显不同。这说明只统计牛之间互动的次数,会掩盖这些互动到底是友好还是竞争性质的信息。

他们做了什么

  1. 研究人员用姿态视觉分析系统处理36头奶牛的监控视频,识别并追踪每头牛,再把两两之间的互动自动分类为舔舐梳理(友好)或驱赶顶撞(敌对)。
  2. 剔除追踪不可靠的记录后,共保留177对牛之间的1,183次互动;人工核查198段样本片段,发现自动分类与人工判断的一致率为82.8%。
  3. 保留下来的互动中72.4%被判定为敌对行为,但互动对象最多的那头牛,并不是网络中充当最多'桥梁'作用(中介中心性最高)的牛。
  4. 按互动性质分开后,友好网络和敌对网络呈现出明显不同的结构:友好网络连接更稀疏但分出的子群体更明显(5个),敌对网络连接更密集但子群体划分较弱(4个)——这种差异在合并统计时完全看不出来。
  5. 作者强调,这些结果只是对一个牛栏区域、一次录像时段的描述性观察,尚不能证明牛群关系稳定,也不能直接当作动物福利评估指标使用,还需要长期验证。
Interaction valence reveals contrasting social networks in dairy cattle figure 0
Interaction valence reveals contrasting social networks in dairy cattle figure 2
Figure 4. Combined undirected interaction network for the 36 observed cows. Node size represents degree, edge width represents cumulative dyadic interaction duration, and node colour represents
Figure 4. Combined undirected interaction network for the 36 observed cows. Node size represents degree, edge width represents cumulative dyadic interaction duration, and node colour represents
Interaction valence reveals contrasting social networks in dairy cattle figure 4

为什么重要

传统的畜牧监控系统往往只统计动物互动的次数,而这项研究表明,把友好接触和竞争性接触混在一起统计,可能掩盖与压力、资源竞争等福利问题密切相关的真实社交结构。对于开发AI动物监测工具的人来说,这提醒我们不仅要检测'是否发生了互动',更要判断'发生的是哪种互动'。

Interaction valence reveals contrasting social networks in dairy cattle figure 5
Interaction valence reveals contrasting social networks in dairy cattle figure 6
Figure 6. Individual variation in observed connectivity, affiliative brokerage, interaction time, and valence composition. Each cow is positioned by degree in the combined network and normalized betweenness in the affiliative network; bubble size represents non-overlapping interaction time and colour represents the predicted agonistic-duration fraction. Labels identify illustrative metric profiles
Figure 6. Individual variation in observed connectivity, affiliative brokerage, interaction time, and valence composition. Each cow is positioned by degree in the combined network and normalized betweenness in the affiliative network; bubble size represents non-overlapping interaction time and colour represents the predicted agonistic-duration fraction. Labels identify illustrative metric profiles
Figure 7. Retained interaction-event counts and predicted agonistic-event proportions across 11 consecutive processing windows. Bars show event counts and the line shows the proportion of retained events classified as agonistic; the horizontal reference denotes the event-weighted session proportion of 72.4%. Window durations differed from 26 to 49 min, and duration-adjusted event rates are reporte
Figure 7. Retained interaction-event counts and predicted agonistic-event proportions across 11 consecutive processing windows. Bars show event counts and the line shows the proportion of retained events classified as agonistic; the horizontal reference denotes the event-weighted session proportion of 72.4%. Window durations differed from 26 to 49 min, and duration-adjusted event rates are reporte

本文术语

  • 姿态视觉追踪 · 通过追踪视频中身体关键点(如头部、四肢)的位置来分析动作和行为的AI方法
  • 二元互动事件 · 记录两个个体(此处为两头牛)之间发生的一次具体互动
  • 友好行为/敌对行为 · 友好行为指梳理等增进亲密的举动,敌对行为指推挤、顶撞等带有竞争或冲突性质的举动
  • 中介中心性 · 衡量一个个体在网络中位于其他个体之间最短连接路径上的频率,体现其'桥梁'作用
  • 模块度(社群检测) · 用于在网络中找出内部联系紧密、与外部联系较少的子群体(社群)的分析方法
Interaction valence reveals contrasting social networks in dairy cattle figure 9
Interaction valence reveals contrasting social networks in dairy cattle figure 10
Figure 8. Within-session interaction-network snapshots. (A) Video 1, 10:30–11:19, 66 active dyads and 66.8% predicted agonistic events; (B) Video 6, 14:20–15:04, 20 active dyads and 85.5% predicted agonistic events; (C) Video 7, 15:04–15:47, 14 active dyads and 53.3% predicted agonistic events; and (D) Video 8, 15:47–16:26, three active dyads and 20.0% predicted agonistic events. Node positions an
Figure 8. Within-session interaction-network snapshots. (A) Video 1, 10:30–11:19, 66 active dyads and 66.8% predicted agonistic events; (B) Video 6, 14:20–15:04, 20 active dyads and 85.5% predicted agonistic events; (C) Video 7, 15:04–15:47, 14 active dyads and 53.3% predicted agonistic events; and (D) Video 8, 15:47–16:26, three active dyads and 20.0% predicted agonistic events. Node positions an

无法转载的图表

  • Figure 2. Computational workflow for constructing valence-aware social networks from video. Cows were detected, tracked and assigned persistent identities; pose trajectories were then used to identify and classify dyadic events. After quality control and production-domain auditing, automated event predictions were aggregated by dyad into weighted, undirected combined, affiliative and agonistic net
在原文中查看图表 →

论文原文摘要（英文）

Social relationships shape access to resources, exposure to conflict and group stability, yet automated livestock monitoring typically treats behaviour as isolated events. Here, we present a valence-aware social-network framework that transforms video-derived interactions into herd-level representations of affiliative and agonistic organization. A pose-based computer-vision pipeline analysed 7 h 39 min of continuous video from the pre-milking area of one commercial dairy farm. After quality control, 1,183 of 1,414 candidate interactions remained, involving 36 cows and 177 dyads. In a predicted-class-balanced audit of 198 pipeline-detected clips, automated and manual labels agreed in 82.8% of cases, with an unweighted audit-sample macro-F1 of 0.872. These values describe the audited sample rather than prevalence-weighted or end-to-end deployment performance. The aggregated network was connected (density = 0.281; transitivity = 0.513; mean path length = 1.88), and predicted affiliative events formed five algorithmic communities (modularity Q = 0.429). Within the observed zone, predicted agonistic interactions comprised 72.4% of retained events and 76.0% of interaction duration. The cow with the most partners did not have the highest betweenness centrality. Separating events by predicted valence produced descriptively different affiliative and agonistic layers, with contrasting edge sets, community partitions and individual positions. Thus, pooled interaction counts can obscure the behavioural composition of an observed network. Valence-aware analysis provides a framework for testing hypotheses about competition, affiliation and welfare-relevant change, while requiring longitudinal validation before use as a welfare or health indicator.

作者 · Sibi Parivendan, Suresh Raja Neethirajan

在 arXiv 阅读

最新论文

全部论文 →

METAL LAB 最新报道

图片来源： Sibi Parivendan et al., arXiv:2608.19222, cc-by-nc-nd-4.0