ALBERT

Description: Endoparasites are one of the major health issues in beef suckler cows and can cause economic losses. As studies on the parasitological status of beef suckler cow herds are rare, the aim of this study was to evaluate the status quo of the parasite burden in herds at four representative locations in Germany. Additionally, the farmers’ pasture management and deworming strategies were documented. Based on these data, the second aim of the study was to develop recommendations for improved deworming and pasture hygiene management. A total of 708 faecal samples were examined with parasitological routine methods. Results revealed Fasciola hepatica, gastrointestinal nematodes (GIN), Eimeria species (spp.), Moniezia spp. and Dictyocaulus viviparus as the most frequent findings. Clinical signs of parasitic diseases were not found during the farm visits. Statistical analyses showed a significant effect of the age status of the animal on the parasitological status in general. Due to the percentage of occurrence, detailed statistical analysis was performed for Eimeria, GIN and Fasciola hepatica, confirming the effect of age status. Assessing the parasitological status of beef suckler cows as routine procedure could help to establish an improved parasite-control management on a farm-individual basis.

Description: Access to adequate foraging material can reduce the occurrence of feather pecking and cannibalism in laying hens. Technical devices may help farmers provide enrichment material more effectively. However, research in this field is rare. On a commercial free-range farm with 15,000 laying hens (Lohmann Tradition), an enrichment device was evaluated from the 30th to the 58th week of age (LW). It ran at five time points (TP) in the afternoon and offered five grams of dried maize silage per hen per day. The numbers of hens residing in defined scratching areas (ScA) either beneath the device (ScA 1 and 3) or in a similar area without the device (ScA 2) were determined. Significantly more hens were found in ScA 1 and ScA 3 when the device was running. On average, only 6.96 (±7.00) hens stayed in ScA 2, whereas 31.45 (±5.38) and 33.83 (±6.16) hens stayed in ScA 1 and ScA 3, respectively. The hen numbers for ScA 1 and ScA 3 did not differ significantly, nor did the TPs have an influence on number of hens within ScA 1 and ScA 3. The number of hens beneath the device can serve as a potential indicator of the device’s usage.

Description: Consistently high data quality is essential for the development of novel loss functions and architectures in the field of deep learning. The existence of such data and labels is usually presumed, while acquiring high-quality datasets is still a major issue in many cases. Subjective annotations by annotators often lead to ambiguous labels in real-world datasets. We propose a data-centric approach to relabel such ambiguous labels instead of implementing the handling of this issue in a neural network. A hard classification is by definition not enough to capture the real-world ambiguity of the data. Therefore, we propose our method “Data-Centric Classification & Clustering (DC3)” which combines semi-supervised classification and clustering. It automatically estimates the ambiguity of an image and performs a classification or clustering depending on that ambiguity. DC3 is general in nature so that it can be used in addition to many Semi-Supervised Learning (SSL) algorithms. On average, our approach yields a 7.6% better F1-Score for classifications and a 7.9% lower inner distance of clusters across multiple evaluated SSL algorithms and datasets. Most importantly, we give a proof-of-concept that the classifications and clusterings from DC3 are beneficial as proposals for the manual refinement of such ambiguous labels. Overall, a combination of SSL with our method DC3 can lead to better handling of ambiguous labels during the annotation process. (Source code is available at https://github.com/Emprime/dc3).

Description: High-quality data is necessary for modern machine learning. However, the acquisition of such data is difficult due to noisy and ambiguous annotations of humans. The aggregation of such annotations to determine the label of an image leads to a lower data quality. We propose a data-centric image classification benchmark with ten real-world datasets and multiple annotations per image to allow researchers to investigate and quantify the impact of such data quality issues. With the benchmark we can study the impact of annotation costs and (semi-)supervised methods on the data quality for image classification by applying a novel methodology to a range of different algorithms and diverse datasets. Our benchmark uses a two-phase approach via a data label improvement method in the first phase and a fixed evaluation model in the second phase. Thereby, we give a measure for the relation between the input labeling effort and the performance of (semi-)supervised algorithms to enable a deeper insight into how labels should be created for effective model training. Across thousands of experiments, we show that one annotation is not enough and that the inclusion of multiple annotations allows for a better approximation of the real underlying class distribution. We identify that hard labels can not capture the ambiguity of the data and this might lead to the common issue of overconfident models. Based on the presented datasets, benchmarked methods, and analysis, we create multiple research opportunities for the future directed at the improvement of label noise estimation approaches, data annotation schemes, realistic (semi-)supervised learning, or more reliable image collection.