Transfer Learning for Object Detection using State-of-the-Art Deep Neural Networks

Talukdar, Jonti; Gupta, Sanchit; Rajpura, Param S.; Hegde, Ravi S.

doi:10.1109/spin.2018.8474198

Cited by 52 publications

(20 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We used transfer learning 12 of a pretrained CNN, to achieve high object detection accuracy on our relatively small data set of 1200 setup device images. ResNet‐50 13 is a convolutional neural network that is trained on more than a million images from the ImageNet database (http://www.image-net.org), and thus has learned rich feature representations for a wide range of images.…”

Section: Methodsmentioning

confidence: 99%

“…ResNet‐50 13 is a convolutional neural network that is trained on more than a million images from the ImageNet database (http://www.image-net.org), and thus has learned rich feature representations for a wide range of images. Networks that are accurate on ImageNet have been shown to be accurate when applied to other image sets using transfer learning, 12 as the networks have learned to extract useful features from natural images that generalize to other data sets. ResNet50 has been trained on over a million images and can classify images into over 1000 object categories, has a depth of 50 layers and contains over 25 million parameters.…”

Section: Methodsmentioning

confidence: 99%

See 1 more Smart Citation

Technical Note: Deep Learning approach for automatic detection and identification of patient positioning devices for radiation therapy

et al. 2020

View full text Add to dashboard Cite

Purpose: Automatic detection and identification of setup devices, using a deep convolutional neural network (CNN) for real-time multiclass object detection, has the potential to reduce errors in the treatment delivery process by avoiding documentation errors. Methods: A database of the setup device photos from the most recent 1200 patients treated at our institution was downloaded from the record and verify (R&V) system along with the corresponding setup notes. Images were manually labeled with bounding boxes of each device. A real-time object detection CNN using the "you only look once" (YOLOv2) architecture was trained using transfer learning of a pretrained CNN (ResNet50). The CNN was trained to detect and identify 11 of the most common treatment accessories used at our institution. Results: Using transfer learning of a CNN for multiclass object detection, we are able to automatically detect and identify setup devices in photographs with an accuracy of 96%. Conclusions: Automation in radiation oncology has the potential to reduce risk. Automatic detection of setup devices is possible using a CNN and transfer learning. This work shows both the value of incident learning systems (ILS) in practice knowledge dissemination, and shows how automation of clinical processes and less reliance on manual documentation has the potential for risk reduction in radiation oncology treatments.

show abstract

Section: Methodsmentioning

confidence: 99%

Section: Methodsmentioning

confidence: 99%

Technical Note: Deep Learning approach for automatic detection and identification of patient positioning devices for radiation therapy

et al. 2020

View full text Add to dashboard Cite

show abstract

“…To overcome these obstacles, a method named transfer learning (Gu et al, 2018) was used. The main objective of this procedure is to transfer the knowledge from one model trained on large datasets, such as ImageNet (Gopalakrishnan et al, 2017), to another model to solve a specific task (Talukdar et al, 2018). Several popular pretrained networks using transfer learning, such as VGG-16, ResNet 50, DeepNet, and AlexNet Inception V2, are described in the literature (Rosebrock, 2018).…”

Section: Fine-tuning and Training Of The Faster-rcnnmentioning

confidence: 99%

A Cloud-Based Environment for Generating Yield Estimation Maps From Apple Orchards Using UAV Imagery and a Deep Learning Technique

et al. 2020

View full text Add to dashboard Cite

Farmers require accurate yield estimates, since they are key to predicting the volume of stock needed at supermarkets and to organizing harvesting operations. In many cases, the yield is visually estimated by the crop producer, but this approach is not accurate or time efficient. This study presents a rapid sensing and yield estimation scheme using offthe-shelf aerial imagery and deep learning. A Region-Convolutional Neural Network was trained to detect and count the number of apple fruit on individual trees located on the orthomosaic built from images taken by the unmanned aerial vehicle (UAV). The results obtained with the proposed approach were compared with apple counts made in situ by an agrotechnician, and an R 2 value of 0.86 was acquired (MAE: 10.35 and RMSE: 13.56). As only parts of the tree fruits were visible in the top-view images, linear regression was used to estimate the number of total apples on each tree. An R 2 value of 0.80 (MAE: 128.56 and RMSE: 130.56) was obtained. With the number of fruits detected and tree coordinates two shapefile using Python script in Google Colab were generated. With the previous information two yield maps were displayed: one with information per tree and another with information per tree row. We are confident that these results will help to maximize the crop producers' outputs via optimized orchard management.

show abstract

“…Various deep neural network architectures have been proposed in past [10], however, our work aimed to use the latest architectures and measure accuracy performance of the networks using small dataset of insects. CNN meta-architectures, namely-Faster-RCNN, SSD Inception and SSD Mobilenet which is obtain very high detection accuracy [11] used in this study.…”

Section: Related Workmentioning

confidence: 99%

Insect Identification Among Deep Learning’s Meta-architectures Using TensorFlow

Patel¹,

Bhatt²

2019

IJEAT

View full text Add to dashboard Cite

Agriculture provides food for human existence, where insects damage the crops. The identification of the insect is a difficult process and subjected to expert opinion. In recent years, researches using deep learning in fields of object detection have been widespread and show accuracy as a result. This study show the comparison of three widely used deep learning meta-architectures (Faster R-CNN, SSD Inception and SSD Mobilenet) as object detection for selected flying insects namely Phyllophaga spp., Helicoverpa armigera and Spodoptera litura. The proposed study is focused on accuracy performance of selected meta-architectures using small dataset of insects. The meta-architecture was tested with same environment for all three architectures and Faster RCNN meta-architecture performs outstanding with an accuracy of 95.33%

show abstract

Transfer Learning for Object Detection using State-of-the-Art Deep Neural Networks

Cited by 52 publications

References 12 publications

Technical Note: Deep Learning approach for automatic detection and identification of patient positioning devices for radiation therapy

Technical Note: Deep Learning approach for automatic detection and identification of patient positioning devices for radiation therapy

A Cloud-Based Environment for Generating Yield Estimation Maps From Apple Orchards Using UAV Imagery and a Deep Learning Technique

Insect Identification Among Deep Learning’s Meta-architectures Using TensorFlow

Contact Info

Product

Resources

About