AI-Driven Yield Estimation Using an Autonomous Robot for Data Acquisition

Mohimont, Lucas; Roesler, Mathias; Steffenel, Luiz Angelo; Gaveau, Nathalie; Rondeau, Marine; Alin, François; Pierlot, Clément; Oliveira, Rachel Ouvinha de; Coppola, Marcello; Doré, Philipe

doi:10.1201/9781003337232-24

Cited by 2 publications

(1 citation statement)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Several deep-learning methods can be used to identify an object in an image. Semantic segmentation is a traditional technique to create a precise mapping of the objects in an image, and was applied to grape counting on [20,21]. However, this technique is computing-expensive and does not scale well for real-time processing.…”

Section: Deep Learning Modelsmentioning

confidence: 99%

Applying Knowledge Distillation on Pre-Trained Model for Early Grapevine Detection

Hollard

Mohimont

2023

Workshop Proceedings of the 19th International Conference on Intelligent Environments (IE2023)

View full text Add to dashboard Cite

The development of Artificial Intelligence has raised interesting opportunities for improved automation in smart agriculture. Smart viticulture is one of the domains that can benefit from Computer-vision tasks through field sustainability. Computer-vision solutions present additional constraints as the amount of data for good training convergence has to be complex enough to cover sufficient features from desired inputs. In this paper, we present a study to implement a grapevine detection improvement for early grapes detection and grape yield prediction whose interest in Champagne and wine companies is undeniable. Earlier yield predictions allow a better market assessment, the harvest work’s organization and help decision-making about plant management. Our goal is to carry estimations 5 to 6 weeks before the harvest. Furthermore, the grapevines growing condition and the large amount of data to process for yield estimation require an embedded device to acquire and compute deep learning inference. Thus, the grapes detection model has to be lightweight enough to run on an embedded device. These models were subsequently pre-trained on two different types of datasets and several layer depth of deep learning models to propose a pseudo-labelling Teacher-Student related Knowledge Distillation. Overall solutions proposed an improvement of 7.56%, 6.98, 8.279%, 7.934% and 13.63% for f1 score, precision, recall, mean average precision at 50 and mean average precision 50-95 respectively on BBCH77 phenological stage.

show abstract

Section: Deep Learning Modelsmentioning

confidence: 99%

Applying Knowledge Distillation on Pre-Trained Model for Early Grapevine Detection

Hollard

Mohimont

2023

Workshop Proceedings of the 19th International Conference on Intelligent Environments (IE2023)

View full text Add to dashboard Cite

show abstract

AI in Current and Future Agriculture: An Introductory Overview

Kisliuk,

Krause,

Meemken

et al. 2023

Künstl Intell

View full text Add to dashboard Cite

In recent years, agriculture has become a major field of application and transfer for AI. The paper gives an overview of the topic, focusing agricultural processes and technology in Central-European style arable farming. AI could also be part of the transformation process of agriculture that is emerging world-wide in response to the UN global sustainable development goals (SDGs). In that sense, our overview includes pointers to some research towards AI in future agricultural systems.

show abstract

AI-Driven Yield Estimation Using an Autonomous Robot for Data Acquisition

Cited by 2 publications

References 0 publications

Applying Knowledge Distillation on Pre-Trained Model for Early Grapevine Detection

Applying Knowledge Distillation on Pre-Trained Model for Early Grapevine Detection

AI in Current and Future Agriculture: An Introductory Overview

Contact Info

Product

Resources

About