Search research articles

ABOUT JoVE

Overview Leadership Blog JoVE Help Center

AUTHORS

Publishing Process Editorial Board Scope & Policies Peer Review FAQ Submit

LIBRARIANS

Testimonials Subscriptions Access Resources Library Advisory Board FAQ

RESEARCH

JoVE Journal Methods Collections JoVE Encyclopedia of Experiments Archive

EDUCATION

JoVE Core JoVE Business JoVE Science Education JoVE Lab Manual Faculty Resource Center Faculty Site

Terms & Conditions of Use

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by

Same author

Molecular diversity and evolutionary dynamics of phage receptor binding proteins in Staphylococcus aureus: implications for phage therapy and resistance.

Scientific reports·2026

Same author

Learning to Diagnose Privately: DP-Powered LLMs for Radiology Report Classification.

IEEE access : practical innovations, open solutions·2026

Same author

Association between household fuel use and cardiometabolic risk factors in sub-Saharan Africa: A systematic review.

Environmental pollution (Barking, Essex : 1987)·2026

Same author

Single-Center Experience with the Optiblock Coil: An Efficient and Thrombogenic Solution for Targeted Vessel Takedown.

World neurosurgery·2026

Same author

Image-based high-throughput phenotyping enables genetic analyses of pod morphological traits in mungbean (Vigna radiata (L.) R. Wilczek).

G3 (Bethesda, Md.)·2026

Same author

Coffee and tea intake and stroke events among West Africans: findings from the SIREN Study.

BMC health services research·2026

Same journal

A transformer-based language model reveals developmental constraint and network complexity during zebrafish embryogenesis.

PNAS nexus·2026

Same journal

Dual phosphoregulatory mechanisms of condensin I revealed by biochemical reconstitution.

PNAS nexus·2026

Same journal

Vanin-1 deficiency enhances host tolerance to influenza infection by modulating cellular redox status.

PNAS nexus·2026

Same journal

Free will in the eyes of Muslims and Christians.

PNAS nexus·2026

Same journal

Paradoxical coexistence of superconductivity and magnetism, and explaining unexpected preferred domain orientations.

PNAS nexus·2026

Same journal

Large language models instantiate evolutionarily robust strategies of cooperation.

PNAS nexus·2026

See all related articles

Search research articles

Related Experiment Video

Updated: May 29, 2025

Flying Insect Detection and Classification with Inexpensive Sensors

Flying Insect Detection and Classification with Inexpensive Sensors

Published on: October 15, 2014

InsectNet: Real-time identification of insects using an end-to-end machine learning pipeline.

Shivani Chiranjeevi¹, Mojdeh Saadati², Zi K Deng³

¹Department of Mechanical Engineering, Iowa State University, Ames, IA 50011, USA.

|February 3, 2025

Summary

This summary is machine-generated.

InsectNet uses deep learning to accurately identify insects, aiding pest management. This AI tool enhances agricultural productivity by distinguishing beneficial and harmful species, even in challenging conditions.

Keywords:

citizen science deep learning insect species classification insect species identification self-supervised learning

More Related Videos

Author Spotlight: AI-Driven Trypanosome Species Detection from Microscopic Images

Author Spotlight: AI-Driven Trypanosome Species Detection from Microscopic Images

Published on: October 27, 2023

Fully Automated Leg Tracking in Freely Moving Insects using Feature Learning Leg Segmentation and Tracking FLLIT

Fully Automated Leg Tracking in Freely Moving Insects using Feature Learning Leg Segmentation and Tracking FLLIT

Published on: April 23, 2020

Related Experiment Videos

Last Updated: May 29, 2025

Flying Insect Detection and Classification with Inexpensive Sensors

Flying Insect Detection and Classification with Inexpensive Sensors

Published on: October 15, 2014

Author Spotlight: AI-Driven Trypanosome Species Detection from Microscopic Images

Author Spotlight: AI-Driven Trypanosome Species Detection from Microscopic Images

Published on: October 27, 2023

Fully Automated Leg Tracking in Freely Moving Insects using Feature Learning Leg Segmentation and Tracking FLLIT

Fully Automated Leg Tracking in Freely Moving Insects using Feature Learning Leg Segmentation and Tracking FLLIT

Published on: April 23, 2020

Area of Science:

Agricultural Science
Computer Science
Entomology

Background:

Insect pests pose significant threats to global agricultural productivity and crop quality.
Accurate insect identification is crucial for effective integrated pest management (IPM).
Automated insect identification faces challenges like species similarity, life stages, and varied imaging conditions.

Purpose of the Study:

To develop an end-to-end deep learning approach, InsectNet, for automated insect identification.
To address challenges in real-world insect identification, including accuracy and trustworthiness.
To create a scalable and accessible AI solution for diverse agricultural ecosystems.

Main Methods:

Utilized a large, citizen-science-sourced insect image dataset with label-free self-supervised learning for global model training.
Fine-tuned the global model with smaller, expert-verified regional datasets for localized identification.
Implemented machine learning operations for streamlined deployment and accessibility.

Main Results:

Achieved >96% accuracy in identifying numerous agriculturally and ecologically relevant insect species.
Demonstrated fine-grained species identification, effective performance in challenging backgrounds, and uncertainty-aware predictions.
Developed a trustworthy and robust insect identification system.

Conclusions:

The global-to-local model strategy offers a scalable and cost-effective solution for insect identification.
InsectNet enhances model trustworthiness and democratizes access to advanced AI for agriculture.
InsectNet is a valuable AI tool complementing existing approaches for agricultural challenges.