nilima kulkarni - Academia.edu (original) (raw)

nilima  kulkarni

Related Authors

Beat Signer

Timothy Jordan

James Elkins

Ellips Masehian

Mark Billinghurst

Woontack Woo

Woontack Woo

Korea Advanced Institute of Science and Technology

Maurizio Forte

David Seamon

Remo Caponi

Fabio Cuzzolin

Uploads

Papers by nilima kulkarni

Research paper thumbnail of Sea Debris Detection Using Deep Learning : Diving Deep into the Sea

2021 IEEE 4th International Conference on Computing, Power and Communication Technologies (GUCON), 2021

Debris produced by humans is typically released into natural underwater environments like rivers ... more Debris produced by humans is typically released into natural underwater environments like rivers and oceans. It's essential to spot marine debris in rivers and oceans so that its impact on the ecosystem can be recognised and reduced. Manually determining the amount of debris present in the oceans is time-consuming, labour-intensive, and has a limited coverage area. This paper aims at a system for the automatic detection and extraction of the debris region in an input image. We have used a deep-learning-based algorithm to perform the task of visually detecting waste in natural underwater ecosystem, with the goal of eventually deploying Autonomous Underwater Vehicles (AUVs) to explore, track, and remove that debris. The deep learning model (InceptionResNetV2 architecture) is used for training. A vast and publicly accessible dataset of real debris in assorted areas is annotated by taking it down from the JAMSTEC e-library of deep-sea debris database. The qualified network is then tested on a series of images from other parts of the dataset, yielding information on how to improve an AUV's detection capability for underwater trash removal. The trained system detects features of interest in a test image using the bounding box and class label maps. The proposed deep learning-based model architecture predicts the class labels with an accuracy of 96% and the bounding box with an accuracy of 82%. Our findings are expected to advance the goals of using AUVs to automatically survey, identify, and capture aquatic debris in underwater environments.

Research paper thumbnail of Color Thresholding Method for Image Segmentation of Natural Images

International Journal of Image, Graphics and Signal Processing, 2012

Research paper thumbnail of Diabetic Retinopathy Stage Classification

SSRN Electronic Journal

Diabetic Retinopathy is the disease of the retina caused by long-standing diabetes leading to dam... more Diabetic Retinopathy is the disease of the retina caused by long-standing diabetes leading to damage of the retina and can even lead to blindness. Diabetic Retinopathy stage classification deals with different stages of diabetes and has been considered as the most important step in the analysis of the disease. Diagnosis of the disease manually requires skilled individuals to identify distinct features, which is a difficult and time-consuming task. In this paper, a deep learning-based CNN model for diagnosing is presented. Digital color fundus images are used. In this work, we have implemented an InceptionV3 model with an attention mechanism. The motive behind using attention mechanism was that the model should pay more attention to the relevant portion of the image. This work will enable a user to test their eyes and can be easily identified in which stage of diabetes they are suffering. All the images were divided into five groups. In the proposed work, the CNN model (InceptionV3) was trained for producing a model for predicting on validation samples. Categorical accuracy of 60.82% is achieved with a Top 2 accuracy of 82.16.

Research paper thumbnail of Smart Transportation Based Car Pooling System

E3S Web of Conferences

Sharing of car is Carpooling, so that multiple number of people can travel in single car. the use... more Sharing of car is Carpooling, so that multiple number of people can travel in single car. the use of car pooling reduces single individuals Travelling cost , reduces the fuel cost and reduces the number of cars. Due to the growth in the population there is inadequate transportation through their own car . Rather than using different mode of Transportation. It results in increasing number of Traffic on roads also increases pollution and increases time to travel to there destination. So by Smart transportation using car pooling system the individual can travel and share there rides with different people of a same destination. In this paper we have carried outs survey. By reviewing various Literature papers on carpooling it aims to reduce the number of cars by sharing the rides.

Research paper thumbnail of Sea Debris Detection Using Deep Learning : Diving Deep into the Sea

2021 IEEE 4th International Conference on Computing, Power and Communication Technologies (GUCON), 2021

Debris produced by humans is typically released into natural underwater environments like rivers ... more Debris produced by humans is typically released into natural underwater environments like rivers and oceans. It's essential to spot marine debris in rivers and oceans so that its impact on the ecosystem can be recognised and reduced. Manually determining the amount of debris present in the oceans is time-consuming, labour-intensive, and has a limited coverage area. This paper aims at a system for the automatic detection and extraction of the debris region in an input image. We have used a deep-learning-based algorithm to perform the task of visually detecting waste in natural underwater ecosystem, with the goal of eventually deploying Autonomous Underwater Vehicles (AUVs) to explore, track, and remove that debris. The deep learning model (InceptionResNetV2 architecture) is used for training. A vast and publicly accessible dataset of real debris in assorted areas is annotated by taking it down from the JAMSTEC e-library of deep-sea debris database. The qualified network is then tested on a series of images from other parts of the dataset, yielding information on how to improve an AUV's detection capability for underwater trash removal. The trained system detects features of interest in a test image using the bounding box and class label maps. The proposed deep learning-based model architecture predicts the class labels with an accuracy of 96% and the bounding box with an accuracy of 82%. Our findings are expected to advance the goals of using AUVs to automatically survey, identify, and capture aquatic debris in underwater environments.

Research paper thumbnail of Color Thresholding Method for Image Segmentation of Natural Images

International Journal of Image, Graphics and Signal Processing, 2012

Research paper thumbnail of Diabetic Retinopathy Stage Classification

SSRN Electronic Journal

Diabetic Retinopathy is the disease of the retina caused by long-standing diabetes leading to dam... more Diabetic Retinopathy is the disease of the retina caused by long-standing diabetes leading to damage of the retina and can even lead to blindness. Diabetic Retinopathy stage classification deals with different stages of diabetes and has been considered as the most important step in the analysis of the disease. Diagnosis of the disease manually requires skilled individuals to identify distinct features, which is a difficult and time-consuming task. In this paper, a deep learning-based CNN model for diagnosing is presented. Digital color fundus images are used. In this work, we have implemented an InceptionV3 model with an attention mechanism. The motive behind using attention mechanism was that the model should pay more attention to the relevant portion of the image. This work will enable a user to test their eyes and can be easily identified in which stage of diabetes they are suffering. All the images were divided into five groups. In the proposed work, the CNN model (InceptionV3) was trained for producing a model for predicting on validation samples. Categorical accuracy of 60.82% is achieved with a Top 2 accuracy of 82.16.

Research paper thumbnail of Smart Transportation Based Car Pooling System

E3S Web of Conferences

Sharing of car is Carpooling, so that multiple number of people can travel in single car. the use... more Sharing of car is Carpooling, so that multiple number of people can travel in single car. the use of car pooling reduces single individuals Travelling cost , reduces the fuel cost and reduces the number of cars. Due to the growth in the population there is inadequate transportation through their own car . Rather than using different mode of Transportation. It results in increasing number of Traffic on roads also increases pollution and increases time to travel to there destination. So by Smart transportation using car pooling system the individual can travel and share there rides with different people of a same destination. In this paper we have carried outs survey. By reviewing various Literature papers on carpooling it aims to reduce the number of cars by sharing the rides.

Log In