filmov
tv
Deep learning based plant phenotyping using Mask RCNN
Показать описание
Part of the ECE 542 Virtual Symposium (Spring 2020)
Our project mainly focuses on detection of phenotypic traits such as leaf and collar count in a plant. The images of plants were manually annotated using MATLAB GUI. The data in was generated in COCO format as a JSON file. The proposed implementation is based on Mask R-CNN. Here, ResNet101 is used as the backbone network and the final model is obtained by training the model for 50 epochs with 500 steps per epoch and a learning rate of 0.001. The model output is obtained as bounding box for object detection, mask, class indicating leaf or collar, confidence score and the number of leaves and collars detected. The mAP values are obtained for different IoU threshold values greater than 0.5 and the loss for validation and training with respect to the number of epochs is reported. mAP value of 0.335 was obtained for the final model for an IoU threshold of 0.5 and the number of leaves and collar detections were compared with the ground truth and found the results to be promising.
Our project mainly focuses on detection of phenotypic traits such as leaf and collar count in a plant. The images of plants were manually annotated using MATLAB GUI. The data in was generated in COCO format as a JSON file. The proposed implementation is based on Mask R-CNN. Here, ResNet101 is used as the backbone network and the final model is obtained by training the model for 50 epochs with 500 steps per epoch and a learning rate of 0.001. The model output is obtained as bounding box for object detection, mask, class indicating leaf or collar, confidence score and the number of leaves and collars detected. The mAP values are obtained for different IoU threshold values greater than 0.5 and the loss for validation and training with respect to the number of epochs is reported. mAP value of 0.335 was obtained for the final model for an IoU threshold of 0.5 and the number of leaves and collar detections were compared with the ground truth and found the results to be promising.
0:08:34
Deep learning based plant phenotyping using Mask RCNN
1:00:15
Deep learning based plant phenotyping with limited labels - Wei Guo | Etincelle#11
0:02:12
Plant Phenotyping by Deep-Learning Based Planner for Multi-Robots
0:59:26
Deep learning for image-based plant phenomics
0:00:47
Image-based Plant Phenotyping
0:23:09
Plant phenotyping with RocketML Deep Learning
1:07:03
Dr. Arti Singh: Image-based plant stress phenotyping in the fast changing world of AI
0:00:48
SMART: Plant phenotyping using Computer Vision and Machine Learning
0:07:24
Plant Phenotyping
0:18:21
Predicting Plant Phenotypes with Deep Learning
0:52:04
2nd IPPN webinar Michael Pound, Nottingham Deep Learning for Plant Phenotyping
0:02:03
Paper Summary - Deep Learning provides SOTA performance in plant phenotyping
0:59:11
Ali Moghimi: Artificial intelligence and hyperspectral imaging for high-throughput plant phenotyping
0:01:10
Plant Phenomics with Deep Learning
0:02:53
Plant Phenotyping with PhenoAIxpert Pro
0:24:51
Tony Pridmore - Image-based Plant Phenotyping - Computer Vision From Lab to Field
0:02:17
AI Driven Pheno Parenting A Deep Learning Based Plant Phenotyping Trait Analysis Model on a Novel So
0:08:00
Plant Phenotyping
0:07:56
Plant Phenotyping using Faster R-CNN
0:37:56
Juan Caicedo - TorBUG Talk - Image-based cell phenotyping with deep learning
3:33:30
Computer Vision in Plant Phenotyping and Agriculture Workshop at ICCV 2021
0:04:51
GRIET MPS 2020-24 Batch -F1 Advanced Plant Phenotyping Using ML Based Image Processing
0:12:28
Applying Machine Learning to Predictive Plant Breeding
0:02:02
ChronoRoot: revealing the secrets of plant root systems using phenotyping and deep learning
Комментарии