Descriptor: Parasitoid Wasps and Associated Hymenoptera Dataset (DAPWH)

Joao Manoel Herrera Pinheiro; Gabriela Do Nascimento Herrera; Luciana Bueno Dos Reis Fernandes; Alvaro Doria Dos Santos; Ricardo V. Godoy; Eduardo A. B. Almeida; Helena Carolina Onody; Marcelo Andrade Da Costa Vieira

cs.CVFeb 20, 2026

Descriptor: Parasitoid Wasps and Associated Hymenoptera Dataset (DAPWH)

Authors: Joao Manoel Herrera Pinheiro, Gabriela Do Nascimento Herrera, Luciana Bueno Dos Reis Fernandes, Alvaro Doria Dos Santos, Ricardo V. Godoy, Eduardo A. B. Almeida, Helena Carolina Onody, Marcelo Andrade Da Costa Vieira,

Organizations: S˜ao Carlos School of Engineering, University of S˜ao Paulo, S˜ao Carlos 13566590, SP, Brazil · Department of Ecology and Evolutionary Biology, Federal University of S˜ao Carlos, S˜ao Carlos 13565905, Brazil · Federal University of Tocantins, Porto Nacional, 77500000, Brazil · Department of Biology (FFCLRP), University of S˜ao Paulo, Ribeir˜ao Preto 14040901, Brazil ·

Abstract

Accurate taxonomic identification is the cornerstone of biodiversity monitoring and agricultural management, particularly for the hyper-diverse superfamily Ichneumonoidea. Comprising the families Ichneumonidae and Braconidae, these parasitoid wasps are ecologically critical for regulating insect populations, yet they remain one of the most taxonomically challenging groups due to their cryptic morphology and vast number of undescribed species. To address the scarcity of robust digital resources for these key groups, we present a curated image dataset designed to advance automated identification systems. The dataset contains 3,556 high-resolution images, primarily focused on Neotropical Ichneumonidae and Braconidae, while also including supplementary families such as Andrenidae, Apidae, Bethylidae, Chrysididae, Colletidae, Halictidae, Megachilidae, Pompilidae, and Vespidae to improve model robustness. Crucially, a subset of 1,739 images is annotated in COCO format, featuring multi-class bounding boxes for the full insect body, wing venation, and scale bars. This resource provides a foundation for developing computer vision models capable of identifying these families.

Figures & tables

Fig. 1: Schematic representation of the dataset file organization. The primary repository (DAPWH) [ 23 ] is structured hierarchically by anatomical view ( Lateral , Frontal , Dorsal_Ventral ), with each view containing specific families’ subdirectories. Complementing this structure, the COCO_subset provides a subset version of the dataset, including the instances_default json file and corresponding images.

Fig. 2: Specimens were retrieved from DCBU collection, mounted under a Leica M205C stereomicroscope equipped with a Leica K5C camera.

Fig. 3: Illustrative geographic distribution of specimen sources from the DCBU, MZUSP, and RPSP collections, generated using Python, Pandas, and Cartopy.

Panel of Figure 4 — Fig. 4: Examples of morphological variation in the DAPWH dataset. (a), (b), (c) Braconidae; (d), (e), (f) Ichneumonidae; (g), (h) Apidae; (i) Vespidae; (j) Bethylidae; (k) Pompilidae; (m) Colletidae.

Panel of Figure 5 — Fig. 5: Examples of morphological variation in the DAPWH dataset. (a), (b), (c) Braconidae; (d), (e), (f) Ichneumonidae; (g), (h) Apidae; (i) Vespidae.

Fig. 6: Illustration of the dataset creation and annotation workflow for the subset COCO annotated dataset. Initial segmentation masks and bounding boxes were generated by computer vision experts in CVAT, assisted by the SAM, and subsequently reviewed by taxonomists.

Family	Dorsal/Ventral	Frontal	Lateral	Total
Ichneumonidae	205	128	453	786
Braconidae	175	66	407	648
Apidae	153	155	158	466
Vespidae	230	87	143	460
Megachilidae	183	19	96	298
Chrysididae	137	31	76	244

TABLE I: Distribution of images per family and view.

\toprule Parameter	Value	Notes
Taxonomic Classes
Ichneumonidae	521	Family level (Masks)
Braconidae	406	Family level (Masks)
Apidae	158	Family level (Masks)
Vespidae	142	Family level (Masks)
Megachilidae	115	Family level (Masks)

TABLE II: Summary of COCO dataset and Image Properties

Model	Acc.	F1	Precision	Recall
EfficientNetV2	92.2%	84.07%	93.69%	82.54%
ConvNeXt Tiny	92.2%	87.00%	92.38%	86.24%
ViTB16	89.3%	80.45%	90.16%	78.48%
VGG16	89.5%	86.39%	88.88%	84.60%
ResNet50	90.1%	84.51%	88.71%	82.10%
YOLOv12	91.7%	95.59%	85.73%	88.88%

TABLE III: Comparison of validation metrics for the evaluated classification models on the test dataset.

Fig. 7: Validation accuracy per epoch across 150 epochs.

Fig. 8: Training and validation metrics over 150 epochs for the YOLOv8 detection model.

Fig. 9: Training and validation metrics over 150 epochs for the YOLOv12 detection model.

Model	mAP50	Precision	Recall
YOLOv8	86.26%	79.08%	85.83%
YOLOv12	90.53%	90.23%	82.54%

TABLE IV: Comparative performance metrics for object detection architectures on the DAPWH test dataset.

Fig. 10: Confusion matrix normalized for YOLOv12.

Panel of Figure 15 — Fig. 11: Visual comparison of detection results. (a,d) The corresponding ground truth annotations; (b,e) The predicted bounding boxes generated by the YOLOv8 during inference; (c,f) The predicted bounding boxes generated by the YOLOv12 during inference.

Biodiversity Monitoring

See all 23 papers

Automated identification of Ichneumonoidea wasps via YOLO-based deep learning: Integrating HiresCam for Explainable AI
Mar 17, 2026Joao Manoel Herrera Pinheiro, Gabriela Do Nascimento Herrera, Alvaro Doria Dos Santos +7Biodiversity Monitoring Image Classification
São Carlos School of Engineering, University of São Paulo, São Carlos 13566590, SP, Brazil · Department of Ecology and Evolutionary Biology, Federal University of São Carlos, São Carlos 13565905, Brazil · Federal University of Tocantins, Porto Nacional, 77500000, Brazil +2
Deep learning-based hierarchical insect classification using camera trap imagery
Jul 30, 2026Zaki Mahfoud, Juan A. Chiavassa, Simon Walther +2Hierarchical Classification Wildlife Monitoring
Weihenstephan-Triesdorf University of Applied Sciences, Department of Sustainable Agriculture and Energy Systems, Professorship of Smart Farming, Am Staudengarten 1, 85354 Freising, Germany · Weihenstephan-Triesdorf University of Applied Sciences, Institute for Biomass Research, Neuseser Str. 1, 91732 Merkendorf, Germany · Weihenstephan-Triesdorf University of Applied Sciences, Department of Sustainable Agriculture and Energy Systems, Professorship of Digital Farm Management, Am Staudengarten 1, 85354 Freising, Germany
eLasmobranc Dataset: An Image Dataset for Elasmobranch Species Recognition and Biodiversity Monitoring
Mar 11, 2026Ismael Beviá-Ballesteros, Mario Jerez-Tallón, Nieves Aranda-Garrido +6Wildlife Monitoring Fine-Grained Image Classification
University of Alicante · Department of Computer Science and Technology (DTIC) · Marine Research Center of Santa Pola (CIMAR)

Mar 17, 2026J/K move · Enter open · S save

Automated identification of Ichneumonoidea wasps via YOLO-based deep learning: Integrating HiresCam for Explainable AI

Joao Manoel Herrera Pinheiro, Gabriela Do Nascimento Herrera, Alvaro Doria Dos Santos, Luciana Bueno Dos Reis Fernandes+6

São Carlos School of Engineering, University of São Paulo, São Carlos 13566590, SP, Brazil · Department of Ecology and Evolutionary Biology, Federal University of São Carlos, São Carlos 13565905, Brazil · Federal University of Tocantins, Porto Nacional, 77500000, Brazil+2

Biodiversity Monitoring Image Classification

Open paper AskPDF arXiv

Accurate taxonomic identification of parasitoid wasps within the superfamily Ichneumonoidea is essential for biodiversity assessment, ecological monitoring, and biological control programs. However, morphological similarity, small body size, and fine-grained interspecific variation make manual identification labor-intensive and expertise-dependent. This study proposes a deep learning-based framework for the automated identification of Ichneumonoidea wasps using a YOLO-based architecture integrated with High-Resolution Class Activation Mapping (HiResCAM) to enhance interpretability. The proposed system simultaneously identifies wasp families from high-resolution images. The dataset comprises 3556 high-resolution images of Hymenoptera specimens. The taxonomic distribution is primarily concentrated among the families Ichneumonidae (n = 786), Braconidae (n = 648), Apidae (n = 466), and Vespidae (n = 460). Extensive experiments were conducted using a curated dataset, with model performance evaluated through precision, recall, F1 score, and accuracy. The results demonstrate high accuracy of over 96 % and robust generalization across morphological variations. HiResCAM visualizations confirm that the model focuses on taxonomically relevant anatomical regions, such as wing venation, antennae segmentation, and metasomal structures, thereby validating the biological plausibility of the learned features. The integration of explainable AI techniques improves transparency and trustworthiness, making the system suitable for entomological research to accelerate biodiversity characterization in an under-described parasitoid superfamily.

Abstract

Biodiversity Monitoring

Explore similar work

Automated identification of Ichneumonoidea wasps via YOLO-based deep learning: Integrating HiresCam for Explainable AI

Deep learning-based hierarchical insect classification using camera trap imagery

eLasmobranc Dataset: An Image Dataset for Elasmobranch Species Recognition and Biodiversity Monitoring