Image caption generator is a task that involves computer vision andnatural language processing concepts to recognize the context of animage and describe them in a natural language like English. The image features will be extracted from Xception which is a CNN model trained on the imagenet dataset and then we feed the features into the LSTM model which will be responsible for generating the image captions.

Finding their way on the streets is one of the many difficulties that blind people face in life. There are a lot of vehicles and obstacles on the street that could get in their way and hurt them. So, keeping this issue in mind, we created a Smart blind stick that uses an ultrasonic sensor to look for obstacles in front of it. We make use of a NodeMCU board and an ultrasonic sensor called the HC-SR04. Complete the circuit then, at that point, transfer the given code to the NodeMCU.

This project performs object detection and tracking using a pre-trained Tensor Flow Lite (TFLite) model. In addition, it can track each unique object in terms of how it is moving through the frame of vision i.e. if it is moving up / down / left / right or just stationary.

"Hummer-Bot" is a multifunctional car based on the BLE-UNO and MX1616L motor. Compared with the traditional car, "Hummer-Bot" is also equipped with wireless control (Bluetooth, infrared, WIFI, and so on). It can trace and avoid obstacles automatically and can also manually control the car with wireless and make full use of each module, as well as integrate with all kinds of related sensors to make the car more intelligent, which is more challenging.

A keyphrase detector, often referred to a keyword spotter (KWS) is a simple speech processing application that detects the presence of a predefined word or short phrase in stream of audio. This is commonly encountered nowadays with hotwords (or wake words) such as "OK Google" or "Alexa" that are used by digital assistants to tell them when to start listening.

The soldier Health and Position Tracking System allows military to track the current GPS position of soldier and also checks the health status including body temperature and heartbeats of soldier. The System also consists extra feature with the help of that soldier can ask for help manually or send a distress signal to military if he is in need. The GPS modem sends the latitude and longitude position with link pattern with the help of that military can track the current position of the soldier. The system is very helpful for getting health status information of soldier and providing them instant help.

Teachable machine allows you to quickly and interactively train a computer vision system to recognize objects simply by offering up examples to a camera and pressing one of 4 buttons to select the classification. The project is a demonstration of the use of pre-trained embeddings for classification which obviates the need for retraining a whole network (potentially at the expense of accuracy).

The pulsation system allows cyclical changes in pressure (vacuum) in the chamber around the liner. This causes the teat cup liner to open allowing milk to flow, then closes the liner again around the teat, massaging the tissues and reducing congestion.

This project uses a Deep Neural Network, more specifically a Convolutional Neural Network, to differentiate between images of people with and without masks. The CNN manages to get an accuracy of 98.2% on the training set and 97.3% on the test set. Then the stored weights of this CNN are used to classify as mask or no mask, in real time, using OpenCV. With the webcam capturing the video, the frames are preprocessed and and fed to the model to accomplish this task. The model works efficiently with no apparent lag time between wearing/removing mask and display of prediction.

The main purpose of this project is, if there will be no traffic on the other signal, one shouldn’t wait for that signal. The system will skip that signal and will move on the next one. Arduino is the main part of this project and it will be used to read from ultrasonic sensor HC-SR04 and calculate the distance. This distance will tell us if any vehicle is near the signal or not and according to that the traffic signals will be controlled.