Tushar Totla's repositories
fire-detection
A artificial intelligence project which detects the fire and smoke before the accident causing huge loses. The work depicts the techniques like object detection, which are in todays world taking great heights. A proposal to a new method based on a deep learning approach, which uses a convolutional neural network that employs dilated convolutions has been approched. The surveillance cameras keep capturing the images, video clips of the region, which then are acted as input to the proposed system and design. It then checks the if the image has been observed with some fire or not and if found rings the alarm and thus the fire management is known in advance and the fire is prevented to avoid any kind of small damage even. In this article, two custom CNN models have been implemented for a cost-effective fire detection CNN architecture for surveillance videos. The first model is a customized basic CNN architecture inspired by AlexNet architecture. We will implement and see its output and limitations and create a customized InceptionV3 model. To balance the efficiency and accuracy, the model is fine-tuned considering the nature of the target problem and fire data. In addition, the method would be designed to be well generalized for unseen data, which offers effective generalization and reduces the number of false alarms.
crypto-stegano-graphy
A secure system which uses both the techniques of steganography and cryptography for secure communication. RSA, AES, LSB algorithms used to achieve the combined system. The plain text goes through cryptographic techniques resulting to cipher text which acts as input to steganographic technique where the output is final cipher. On reciever side the reverse process is done to decrypt or say decode the message. The system bring double security to the message transferring An additional feature of mail is even embedded into the system where usage of SMTP is done.to
sentiment-analysis
The project or work which goals to extract the opinions, emotions, attitudes of public towards different object of interest. Sentiment analysis is a form of shallow semantic analysis of texts. In the project an automatic approach that involves supervised machine learning and text mining classification algorithms are used which includes the sentimental analysis in various applications. Various fields like twitter tweets, movie review tweets, election result tweets, digital libraries, life sciences, social media tweets and various other fields have been analyzed and the algorithms like line regression, Support vector machine (SVM), Naïve bayes are used in every content and a result is brought out through the means of various types of graphs. More specifically, MOVIE REVIEWS have been considered and the scrapping reviews of a particular movie after the predicting the sentiment of each review and on that basis, status of the movie review is understood. It is predicted that one can generalize the other applications using the specific content even.
password_manager
A chrome extension to prevent the attacks like cross site scripting and to authenticate the user before enabling the auto fill option for logins, forms. An user friendly interface chromium extension for avoiding various cyber attacks.