Lightweight SAR Ship detection and 16 Class Classification using Novel Deep Learning Algorithm with a Hybrid Preprocessing Technique

2022 ◽  
pp. 1-28
Author(s):  
Anil Raj J ◽  
Sumam Mary Idicula ◽  
Binu Paul
Keyword(s):  
Deep Learning ◽  
Learning Algorithm ◽  
Ship Detection ◽  
Deep Learning Algorithm ◽  
Preprocessing Technique

scholarly journals Kathakali face expression detection using deep learning techniques

2021 ◽  
Vol 2062 (1) ◽  
pp. 012018
Author(s):  
C Selvi ◽  
Y Anvitha ◽  
C H Asritha ◽  
P B Sayannah
Keyword(s):  
Deep Learning ◽  
Facial Expressions ◽  
Learning Algorithm ◽  
Paper Machine ◽  
Classical Dance ◽  
Face Expression ◽  
Deep Learning Algorithm ◽  
Art Form ◽  
Preprocessing Technique ◽  
Processing Techniques

Abstract To develop a Deep Learning algorithm that detects the Kathakali face expression (or Navarasas) from a given image of a person who performs Kathakali. One of India’s major classical dance forms is Kathakali. It is a “story play” genre of art, but one distinguished by the traditional male-actor-dancers costumes, face masks and makeup they wear. In the Southern region of India, Kathakali is a Hindu performance art in Malayalam speaking. Most of the plays are epic scenes of Mahabharata and Ramayana. A lot of foreigners visiting India are inspired by this art form and have been curious about the culture. It is still used for entertainment as a part of tourism and temple rituals. An understanding of facial expressions are essential so as to enjoy the play. The scope of the paper is to identify the facial expressions of Kathakali to have a better understanding of the art play. In this paper, Machine Learning and Image Processing techniques are used to decode the expressions. Kathakali face expressions are nine types namely-Adbhutam (wonder), Hasyam (comic), Sringaram(love), Bheebatsam(repulsion), Bhayanakam(fear), Roudram(anger), Veeram(pride), Karunam(sympathy) and Shantham (peace). These Expressions are mapped to real world human emotions for better classification through face detection and extraction to achieve the same. Similarly a lot of research in terms of Preprocessing and Classification is done to achieve the maximum accuracy. Using CNN algorithm 90% of the accuracy was achieved. In order to conserve the pixel distribution and as no preprocessing was used for better object recognition and analysis Fuzzy algorithm is taken into consideration. Using this preprocessing technique 93% accuracy was achieved.

scholarly journals Study on Radar Echo-Filling in an Occlusion Area by a Deep Learning Algorithm

2021 ◽  
Vol 13 (9) ◽  
pp. 1779
Author(s):  
Xiaoyan Yin ◽  
Zhiqun Hu ◽  
Jiafeng Zheng ◽  
Boyong Li ◽  
Yuanyuan Zuo
Keyword(s):  
Deep Learning ◽  
Loss Function ◽  
Learning Algorithm ◽  
Weather Radar ◽  
Loss Functions ◽  
Training Dataset ◽  
Echo Intensity ◽  
Common Mean ◽  
Deep Learning Algorithm ◽  
Radar Beam

Radar beam blockage is an important error source that affects the quality of weather radar data. An echo-filling network (EFnet) is proposed based on a deep learning algorithm to correct the echo intensity under the occlusion area in the Nanjing S-band new-generation weather radar (CINRAD/SA). The training dataset is constructed by the labels, which are the echo intensity at the 0.5° elevation in the unblocked area, and by the input features, which are the intensity in the cube including multiple elevations and gates corresponding to the location of bottom labels. Two loss functions are applied to compile the network: one is the common mean square error (MSE), and the other is a self-defined loss function that increases the weight of strong echoes. Considering that the radar beam broadens with distance and height, the 0.5° elevation scan is divided into six range bands every 25 km to train different models. The models are evaluated by three indicators: explained variance (EVar), mean absolute error (MAE), and correlation coefficient (CC). Two cases are demonstrated to compare the effect of the echo-filling model by different loss functions. The results suggest that EFnet can effectively correct the echo reflectivity and improve the data quality in the occlusion area, and there are better results for strong echoes when the self-defined loss function is used.

scholarly journals An Application of Deep Learning Algorithm for Automatic Multiclass Instance Segmentation of Concrete Damage

IEEE Access ◽  
2021 ◽  
pp. 1-1
Author(s):  
Prashant Kumar ◽  
Apurva Sharma ◽  
Solomon Raju Kota
Keyword(s):  
Deep Learning ◽  
Learning Algorithm ◽  
Deep Learning Algorithm ◽  
Concrete Damage ◽  
Instance Segmentation
Sign in / Sign up

Export Citation Format

Share Document