Description:

Plant disease has long been one of the major threats to food security because it drastically reduces the crop yield and compromises its quality. Deep learning is typically used as it allows the computer to autonomously learn the most suitable feature without human intervention. The system is mainly focused on finding the whether a plant is attacked by any disease or not. CNN (convolutional neural network) is widely used among in image recognition among various network architectures in deep learning

Name(s):D. Srinivas (17H61A0511),E. Abhishek Gupta(17H61A0515),Y. Rakesh

Description:

Due to varying extreme climatic conditions like in the Philippines, Filipinos tend to develop varied skin conditions. Skin diseases are characterized as disorders that frequently begin interior the body or start from the skin and clearly show up on the skin. It is important to determine the type of skin condition as soon as possible. This can sometimes become a dangerous cancer. If cancer is diagnosed, treatment must be started immediately, otherwise it may be life-threatening. If this type of disease is determined earlier there may be a chance of curability. A model was developed to detect the skin diseases by using Convolutional Neural Networks.

Name(s):M Sai Chaitanya (17H61A0530),N Preetha (17H61A0532),T Sagar (18H65A0505)

Description:

Image captioning is a process of generating image description for understanding various elements in the image. Those elements include the object/person in the image, background in the image and relationship between the objects. So, in order to get the image caption, a model is created with CNN and LSTM.

Name(s):G. Sreeja (17H61A0582) ,V. Aravind (17H61A05B2) , Mohd Arif (17H61A0594)

Description:

The most affected cancer in women is breast cancer and the cases of breast cancer have been drastically increasing all around the world in every country. In many the presence of breast cancer is being recognized in the last stages which are affecting the survival rate. In order to classify a biopsy report as positive for malignant lumps the proposed system uses Deep Convolutional Neural Networks. The use of Transfer learning techniques along with the CNNs allows training the model for highest degree of accuracy. This model is implemented through an android app which runs locally on the mobile for easier accessibility. In this, the user uploads the biopsy report through the android application. The report is then fed to the model for classification. The model consists of following layers which are MobileNetV2, Global average pooling layer2d, fully connect layer and output layer. After classification the result is obtained displayed to user through the application.

Name(s):K P NISHIKANT (17H61A05L1),R SHARANI (17H61A05N6),M NIKITHA (17H61A05M3)

Description:

Video Classification is the task of producing a label that is relevant to the video given its frames. A good video level classifier is one that not only provides accurate frame labels, but also best describes the entire video given the features and the annotations of the various frames in the video. It is one of the active research areas in computer vision for various contexts like security surveillance, healthcare and human computer interaction (HCI) and are developed as part of a framework to enable continuous monitoring of human behaviors in the area of ambient assisted living, sports injury detection, elderly care, rehabilitation, and entertainment and surveillance in smart home environment.

Name(s):P. Sathwika Reddy(17H61A05F8) ,K. Sravani Reddy(17H61A05E1), Pavan Raj(17H61A05D0)