Scene Recognition using Places Dataset
Sharmila Agnal. A1, Jenny Savani2, Suchhanda Das3, Preetam Swaraj4, Ayushi Rathi5
1Sharmila Agnal.A, Department of Computer Science & Engineering, SRM Institute of Science and Technology, Chennai, Country, India.
2Jenny Savani, Department of Computer Science & Engineering, SRM Institute of Science and Technology, Chennai, Country, India.
3Suchhanda Das, Department of Computer Science & Engineering, SRM Institute of Science and Technology, Chennai, Country, India.
4Preetam Swaraj, Department of Computer Science & Engineering, SRM Institute of Science and Technology, Chennai, Country, India.
5Ayushi Rathi, Department of Computer Science & Engineering, SRM Institute of Science and Technology, Chennai, Country, India.
Manuscript received on 17 April 2019 | Revised Manuscript received on 24 April 2019 | Manuscript published on 30 April 2019 | PP: 633-636 | Volume-8 Issue-6, April 2019 | Retrieval Number: E3155038519/19©BEIESP
Open Access | Ethics and Policies | Cite | Mendeley | Indexing and Abstracting
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC-BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Abstract: Artificial Intelligence is the major breakthrough in the field of computer science. Moreover, with the help of computer vision and image processing, analyzing and understanding digital images has become very easy. In this paper, we are constructing an Artificial Intelligence model which will detect the scene with maximum accuracy. At first, the machine is trained with various categories of landscapes with each and every image from the collection of labeled pictures from database. Training includes dividing the image into fine sub-regions and analyzing histograms using external and internal features present inside each sub-region. For this, we are using Places database, which is collection of millions of scenic images. These images will be labeled with semantic categories which comprises of a large and diverse list of the types of landscapes found in the world. Also, by using CNN (Convolutional Neural Networks), we learn deep features for scene recognition tasks, and establish several scenic-centric categories. Scene recognition provides simplistic visual sense which helps in understanding relationship between foreground and background in an image. Lastly, we intend to design a system which can build the proposed model using a database consisting of all the possible categories of landscapes.
Keyword: Artificial Intelligence, deep features, image processing, scene recognition
Scope of the Article: Artificial Intelligence