Analyzing Data Modalities for Cattle Weight Estimation Using Deep Learning Models.

We investigate the impact of different data modalities for cattle weight estimation. For this purpose, we collect and present our own cattle dataset representing the data modalities: RGB, depth, combined RGB and depth, segmentation, and combined segmentation and depth information. We explore a recent vision-transformer-based zero-shot model proposed by Meta AI Research for producing the segmentation data modality and for extracting the cattle-only region from the images.

Impact of conservation measures on demography and genetic variability of livestock breeds.

Conservation of animal genetic resources requires regular monitoring and interventions to maintain population size and manage genetic variability. This study uses genealogical information to evaluate the impact of conservation measures in Europe, using (i) data from the Domestic Animal Diversity Information System (DAD-IS) and (ii) a posteriori assessment of the impact of various conservation measures on the genetic variability of 17 at-risk breeds with a wide range of interventions. Analysis of data from DAD-IS showed that 68% of national breed populations reported to receive financial support showed increasing demographic trends, v.

Genetic conditions of joint Nordic genetic evaluations of lifetime competition performance in warmblood sport horses.

Breeding programmes for warmblood sport horses are similar in the Nordic countries Sweden, Denmark, Finland and Norway, and stallions of same origin are used. The aim was to investigate whether a joint Nordic genetic evaluation based on lifetime competition performance is feasible and beneficial for breeding competitive sport horses in the Nordic countries. Results for almost 45,000 horses in show jumping and 30,000 horses in dressage were available.