Benchmark database for fine-grained image classification of benthic macroinvertebrates

Raitoharju, Jenni; Riabchenko, Ekaterina; Ahmad, Iftikhar; Iosifidis, Alexandros; Gabbouj, Moncef; Kıranyaz, Serkan; Tirronen, Ville; Ärje, Johanna; Kärkkäinen, Salme; Meissner, Kristian

doi:10.1016/j.imavis.2018.06.005

Cited by 33 publications

(53 citation statements)

References 36 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…So far, costs are in the same order of magnitude as for as traditional lab and identification procedures [7]. And if abundance estimates are continued to be obtained via traditional inspection or alternatively automated image recognition [40], no substantial cut in costs can be expected. However, the central incentive for including also genetic data should be the fundamentally improved resolution down to species or even population level [41] that can be obtained in a standardised fashion.…”

Section: Discussionmentioning

confidence: 99%

Analysis of 13,312 benthic invertebrate samples from German streams reveals minor deviations in ecological status class between abundance and presence/absence data

et al. 2019

View full text Add to dashboard Cite

Benthic invertebrates are the most commonly used organisms used to assess ecological status as required by the EU Water Framework Directive (WFD). For WFD-compliant assessments, benthic invertebrate communities are sampled, identified and counted. Taxa × abundance matrices are used to calculate indices and the resulting scores are compared to reference values to determine the ecological status class. DNA-based tools, such as DNA metabarcoding, provide a new and precise method for species identification but cannot deliver robust abundance data. To evaluate the applicability of DNA-based tools to ecological status assessment, we evaluated whether the results derived from presence/absence data are comparable to those derived from abundance data. We analysed benthic invertebrate community data obtained from 13,312 WFD assessments of German streams. Broken down to 30 official stream types, we compared assessment results based on abundance and presence/absence data for the assessment modules “organic pollution” (i.e., the saprobic index) and “general degradation” (a multimetric index) as well as their underlying metrics.In 76.6% of cases, the ecological status class did not change after transforming abundance data to presence/absence data. In 12% of cases, the status class was reduced by one (e.g., from good to moderate), and in 11.2% of cases, the class increased by one. In only 0.2% of cases, the status shifted by two classes. Systematic stream type-specific deviations were found and differences between abundance and presence/absence data were most prominent for stream types where abundance information contributed directly to one or several metrics of the general degradation module. For a single stream type, these deviations led to a systematic shift in status from ‘good’ to ‘moderate’ (n = 201; with only n = 3 increasing). The systematic decrease in scores was observed, even when considering simulated confidence intervals for abundance data. Our analysis suggests that presence/absence data can yield similar assessment results to those for abundance-based data, despite type-specific deviations. For most metrics, it should be possible to intercalibrate the two data types without substantial efforts. Thus, benthic invertebrate taxon lists generated by standardised DNA-based methods should be further considered as a complementary approach.

show abstract

Section: Discussionmentioning

confidence: 99%

Analysis of 13,312 benthic invertebrate samples from German streams reveals minor deviations in ecological status class between abundance and presence/absence data

et al. 2019

View full text Add to dashboard Cite

show abstract

“…These methods are all used to learn features that can more robustly distinguish the smaller classes. Species identification performance can vary widely, ranging from species that are correctly identified in most cases to species that are generally difficult to identify (93). Typically, the amount of training data is a key element for successful identification, although recent analyses of images of ∼65,000 specimens in the carabid beetle collection at the Natural History Museum London suggest that imbalances in identification performance are not necessarily related to how well represented a species is in the training data (84).…”

Section: Potential Deep Learning Applications In Entomologymentioning

confidence: 99%

Deep learning and computer vision will transform entomology

Høye

Ärje

Bjerge

et al. 2021

Proc. Natl. Acad. Sci. U.S.A.

Self Cite

304

206

View full text Add to dashboard Cite

Most animal species on Earth are insects, and recent reports suggest that their abundance is in drastic decline. Although these reports come from a wide range of insect taxa and regions, the evidence to assess the extent of the phenomenon is sparse. Insect populations are challenging to study, and most monitoring methods are labor intensive and inefficient. Advances in computer vision and deep learning provide potential new solutions to this global challenge. Cameras and other sensors can effectively, continuously, and noninvasively perform entomological observations throughout diurnal and seasonal cycles. The physical appearance of specimens can also be captured by automated imaging in the laboratory. When trained on these data, deep learning models can provide estimates of insect abundance, biomass, and diversity. Further, deep learning models can quantify variation in phenotypic traits, behavior, and interactions. Here, we connect recent developments in deep learning and computer vision to the urgent demand for more cost-efficient monitoring of insects and other invertebrates. We present examples of sensor-based monitoring of insects. We show how deep learning tools can be applied to exceptionally large datasets to derive ecological information and discuss the challenges that lie ahead for the implementation of such solutions in entomology. We identify four focal areas, which will facilitate this transformation: 1) validation of image-based taxonomic identification; 2) generation of sufficient training data; 3) development of public, curated reference databases; and 4) solutions to integrate deep learning and molecular tools.

show abstract

“…To facilitate the automation of specimen identification, biomass estimation and sorting of invertebrate specimens, we improved the prototype imaging system developed for automatic identication of benthic macroinvertebrates (Raitoharju et al, 2018). We named the new device BIODISCOVER machine, as an acronym for BIOlogical specimens Described, Identified, Sorted, Counted and Observed using Vision-Enabled Robotics.…”

Section: The Biodiscover Machinementioning

confidence: 99%

“…To overcome those limitations, Zhang, Gao, and Caelli (2010) have proposed a method for structuring 3D insect models from 2D images. Raitoharju et al (2018) Raitoharju et al (2018) using industry components to make it completely reproducible. It has been made light proof to prevent extraneous light from affecting the images.…”

Section: Introductionmentioning

confidence: 99%

Automatic image‐based identification and biomass estimation of invertebrates

et al. 2020

Self Cite

View full text Add to dashboard Cite

show abstract

Benchmark database for fine-grained image classification of benthic macroinvertebrates

Abstract: This is a self-archived version of an original article. This version may differ from the original in pagination and typographic details.

Cited by 33 publications

References 36 publications

Analysis of 13,312 benthic invertebrate samples from German streams reveals minor deviations in ecological status class between abundance and presence/absence data

Analysis of 13,312 benthic invertebrate samples from German streams reveals minor deviations in ecological status class between abundance and presence/absence data

Deep learning and computer vision will transform entomology

Automatic image‐based identification and biomass estimation of invertebrates

Contact Info

Product

Resources

About