Skip to content

This project is an attempt at developing an object detection model using modern computer vision technology. The project aims to incorporate state-of-the-art techniques for object detection with the goal of achieving high accuracy in detecting objects. The model is build using the MXNet Scala Inference API or Tensorflow for scala.

License

Notifications You must be signed in to change notification settings

ManaliSharma/BigData_Scala_Cat_Dog_Duck

Repository files navigation

BigData_Scala_Cat_Dog_Duck

Image Classification and Object Detection- Tensorflow , Spark and Scala

Overview

This project is an attempt at developing an object detection model using modern computer vision technology. The project aims to incorporate state-of-the-art techniques for object detection with the goal of achieving high accuracy in detecting objects. The model is build using the MXNet Scala Inference API or Tensorflow for scala.

Dataset being worked upon can be downloaded from here :- https://www.kaggle.com/c/imagenet-object-localization-challenge/data

  • Input: Images
  • Output: Identified Objects (Persian cat, mountain bike, hot dog, etc.) maximum 4 objects will be detected in this image
  • Dataset: ImageNet Dataset (Kaggle)

(for object detection)

Requirements (Libraries)

  • Tensorflow = "0.2.4"
  • BetterFiles = "3.4.0"
  • DL4J = "1.0.0-alpha"
  • Janino = "2.6.1"
  • LogbackClassic = "1.2.3"
  • ScalaCheck = "1.13.5"
  • ScalaTest = "3.0.4"

Folder Structure

  • BigData_Scala_Cat_Dog_Duck
  • README.md
  • Object_Detector
    • Tensorflow
    • src
    • target
    • model
    • Target
  • Classification
    • src
  • Object Detection Model
    • src
    • target
    • .pom files
  • Spark_trial_S3
  • Final_Project.ppt

How to Run

  • Clone the repository in your local disk.
  • Refer the folder structure mentioned above, and open the folder "BigData_Scala_Cat_Dog_Duck".
  • This is the master folder and it links to all the different parts of the project.
  • Please make sure that the libraries mentioned above are all installed
  • Please make sure you have an active AWS account for running the clusters -There should be active spark cluster running

spark-submit --deploy-mode cluster --master yarn --num-executors 6 --conf "spark.serializer=org.apache.spark.serializer.JavaSerializer" --driver-cores 3 --executor-memory 10G --driver-memory 10G --conf spark.yarn.executor.memoryOverhead=2048 --class ImageClassifier Image_Classifier.jar --resolution 300 --model /home/hadoop/analytics-zoo_ssd-vgg16-300x300_PASCAL_0.1.0.model

java -cp tensorflow_2.12-0.1.0-SNAPSHOT.jar input.project.example.ObjectDetector image child.JPEG

About

This project is an attempt at developing an object detection model using modern computer vision technology. The project aims to incorporate state-of-the-art techniques for object detection with the goal of achieving high accuracy in detecting objects. The model is build using the MXNet Scala Inference API or Tensorflow for scala.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages