Edge intelligence for poultry welfare: Utilizing tiny machine learning neural network processors for vocalization analysis

PLoS ONE, Journal Year: 2025, Volume and Issue: 20(1), P. e0316920 - e0316920

Published: Jan. 16, 2025

The health of poultry flock is crucial in sustainable farming. Recent advances machine learning and speech analysis have opened up opportunities for real-time monitoring the behavior flock. However, there has been little research on using Tiny Machine Learning (Tiny ML) continuous vocalization poultry. This study addresses this gap by developing deploying ML models low-power edge devices to monitor chicken vocalizations. focus overcoming challenges such as memory limitations, processing power, battery life ensure practical implementation agricultural settings. In collaboration with avian researchers, a diverse dataset vocalizations representing range environmental conditions was created train validate algorithms. Digital Signal Processing (DSP) blocks Edge Impulse platform were used generate spectral features studying fowl vocalization. A one-dimensional Convolutional Neural Network (CNN) model employed classification. emphasizes accurately identifying categorizing different noises associated emotional states discomfort, hunger, satisfaction. To improve accuracy reduce background noise, noise-robust algorithms developed. Before removal our average F1 scores 91.6% 0.92, respectively. After removal, they improved 96.6% 0.95.

Language: Английский

---

Adapting a Large-Scale Transformer Model to Decode Chicken Vocalizations: A Non-Invasive AI Approach to Poultry Welfare

AI, Journal Year: 2025, Volume and Issue: 6(4), P. 65 - 65

Published: March 25, 2025

Natural Language Processing (NLP) and advanced acoustic analysis have opened new avenues in animal welfare research by decoding the vocal signals of farm animals. This study explored feasibility adapting a large-scale Transformer-based model, OpenAI’s Whisper, originally developed for human speech recognition, to decode chicken vocalizations. Our primary objective was determine whether Whisper could effectively identify patterns associated with emotional physiological states poultry, thereby enabling real-time, non-invasive assessments. To achieve this, data were recorded under diverse experimental conditions, including healthy versus unhealthy birds, pre-stress post-stress scenarios, quiet noisy environments. The audio recordings processed through producing text-like outputs. Although these outputs did not represent literal translations vocalizations into language, they exhibited consistent token sequences sentiment indicators strongly correlated recognized poultry stressors conditions. Sentiment using standard NLP tools (e.g., polarity scoring) identified notable shifts “negative” “positive” scores that corresponded closely documented changes intensity stress events altered states. Despite inherent domain mismatch—given Whisper’s original training on speech—the findings clearly demonstrate model’s capability reliably capture features significant welfare. Recognizing limitations applying English-oriented tools, this proposes future multimodal validation frameworks incorporating sensors behavioral observations further strengthen biological interpretability. our knowledge, work provides first demonstration architectures, even without species-specific fine-tuning, can encode meaningful from vocalizations, highlighting their transformative potential advancing productivity, sustainability, practices precision farming.

Language: Английский

---

Human-computer interactions with farm animals—enhancing welfare through precision livestock farming and artificial intelligence

Frontiers in Veterinary Science, Journal Year: 2024, Volume and Issue: 11

Published: Nov. 14, 2024

While user-centered design approaches stemming from the human-computer interaction (HCI) field have notably improved welfare of companion, service, and zoo animals, their application in farm animal settings remains limited. This shortfall has catalyzed emergence animal-computer (ACI), a discipline extending technology’s reach to multispecies user base involving both animals humans. Despite significant strides other sectors, adaptation HCI ACI (collectively HACI) welfare—particularly for dairy cows, swine, poultry—lags behind. Our paper explores potential HACI within precision livestock farming (PLF) artificial intelligence (AI) enhance individual address unique challenges these settings. It underscores necessity transitioning productivity-focused animal-centered methods, advocating paradigm shift that emphasizes as integral sustainable practices. Emphasizing ‘One Welfare’ approach, this discussion highlights how integrating technologies not only benefits health, productivity, overall well-being but also aligns with broader societal, environmental, economic benefits, considering pressures farmers face. perspective is based on insights one-day workshop held June 24, 2024, which focused advancing welfare.

Language: Английский

---

The effect of worker clothing color on stress in laying hens

Archives animal breeding/Archiv für Tierzucht, Journal Year: 2024, Volume and Issue: 67(2), P. 145 - 151

Published: April 5, 2024

Abstract. The environment in which animals are kept must provide suitable conditions for their species. This includes ensuring that healthy, well-fed, safe, able to exhibit species-specific behaviors, not experiencing fear or pain, and under chronic acute stress. Poultry welfare is achieved when birds raised environments meet physiological ethological needs. Fear can significantly impact animal welfare. Chickens have been altered by human artificial selection. Despite this, they reactivity towards humans tend avoid them. reared environmentally controlled poultry houses bred superior productivity more sensitive factors lost adaptability a great extent. study aimed determine the effect of personnel clothing color on stress chickens layer hen coops. experiment involved 32-week-old laying hens three different genotypes. A worker henhouse wore six respective colors workwear (dark blue, green, red, yellow, black, white), sound measurements were taken during this time. results showed worker's influenced intensity (P<0.05). White elicited least reaction, whereas black dark blue most. other similar reactions. In conclusion, workers coops wearing clothing, such as induce noise animals. Additionally, reactions yellow colors, with white being around felt most secure.

Language: Английский

---

Leveraging Geospatial Technologies for Resource Optimization in Livestock Management

Journal of Geoscience and Environment Protection, Journal Year: 2024, Volume and Issue: 12(10), P. 287 - 307

Published: Jan. 1, 2024

Language: Английский

---