Name
Abstract | ||
|---|---|---|
Over the past decade, entomologists have observed a worldwide decline in the population of pollinating insects [16] , [13] . In Germany, the insect biomass dropped by up to 76.7 % in some areas, including wild bees from 1989 until 2016 [13] . Even though there is a number of hypotheses about of potential underlying causes for this decline, there are concerns that insecticides are at least partially responsible [8] , [17] , [12] . The design of insecticides requires accurate risk assessment procedures to avoid damage to beneficial insects and like pollinators such as the buff-tailed bumblebee ( Bombus terrestris ). In order to provide legal authorities with an accurate risk assessment of plant protection products on bees ( Apis mellifera, Bombus terrestris and solitary bees ) the OECD [15] , EFSA [5] , EPA [6] and other legal entities around the world published protocols, guidance documents and guidelines for such assessment. However, recent studies urge the importance of expanding them to include assessments of effects which are not lethal yet harmful. These sub-lethal effect include the change in be-havioral performance, such as ability of foragers to return to the colony [19] . Current guideline drafts by the EFSA [4] include homing flight studies to test insect behavioral performance. They inspect whether and which individuals return, therefore insect re-ID (re-identification) is crucial [18] . Current bumblebee re-ID techniques are limited to placing markers on individual subjects, which can be analog [11] or digital (RFID) [21] . These steps are very time consuming, intrusive, and are prone to mechanical failure. To find a solution to demedy these shortcomings, a visual approach is proposed. Recently deep learning systems emerged that can learn, through their exposure to many examples, the particular features that allow the discrimination of individuals. Especially the success of re-ID of fruit flies deems the re-ID of bumblebees on images feasible [20] . In order to build such a re-ID system, a dataset is needed. Therefore we propose the following bumblebee re-ID dataset. |
Year | DOI | Venue |
|---|---|---|
2020 | 10.1109/WACVW50321.2020.9096909 | 2020 IEEE Winter Applications of Computer Vision Workshops (WACVW) |
Keywords | DocType | ISSN |
Bumblebee Re-Identification Dataset,entomologists,pollinating insects,Germany,insect biomass,wild bees,insecticides,accurate risk assessment procedures,beneficial insects,pollinators,buff-tailed bumblebee,Bombus terrestris,legal authorities,plant protection products,Apis mellifera,solitary bees,OECD,EPA,legal entities,world published protocols,guidance documents,sub-lethal effect,behavioral performance,current guideline drafts,EFSA,flight studies,insect behavioral performance,insect re-ID,current bumblebee re-ID techniques,individual subjects,deep learning systems,bumblebees,bumblebee re-ID dataset | Conference | 2572-4398 |
ISBN | Citations | PageRank |
978-1-7281-7163-0 | 0 | 0.34 |
References | Authors | |
0 | 4 | |
Name | Order | Citations | PageRank |
|---|---|---|---|
| Frederic Tausch | 1 | 0 | 0.34 |
| Simon Stock | 2 | 0 | 0.68 |
| Julian Fricke | 3 | 0 | 0.34 |
| Olaf Klein | 4 | 0 | 0.34 |