In the project we are trying to implement an app through which was can move around an object in real life and see the various details in it from various perspectives as if the object was present in real word. In order to implement this we are using a popular game engine named “Unity”. We create a plane and project the scene captured by the smartphone’s back camera or “the real world” over it and use the smart phones gyroscope to control the rotation around it. We then make the object as the pivoting point for the main camera to rotate around so that we can rotate around the object and see the various details of it from various perspective and also we can translate or open that app in any place and can see the same object from various position in that place.
We understand that many people wish to have a healthier diet and track what they are eating but find it very cumbersome to do so (imagine manual data entry on an app). Therefore, we want to build a practical and functional spoon that can do all that for us automatically and in a fun way. The implementation of this spoon required the following fields of knowledge: embedded systems (sensors & actuators), cloud computing for the server, data visualization for the dashboard and data science/deep learning for image recognition.
Patched Face Completion using Denoising Autoencoder and GANs
Dataset: CelebA dataset
We take Kolkata/images in batches (here 64), perform augmentation (zooming in/out, shifting across width/height, changing brightness, horizontal flipping etc) over them and then apply a Gaussian Noise and introduce patches of random dimensions in random regions (except at the borders) across the image. This patched image is then fed into a Denoising Autoencoder which acts as the Generator. This generator then tries to reconstruct the patched part of the image. There is also a Discriminator, whose job is to differentiate between the actual Kolkata/images from the dataset itself and the Kolkata/images generated by the generator. The Discriminator should output 1 when the image is from the actual dataset else it should output 0. The discriminator uses a binary cross entropy loss for its loss function The generator uses a binary cross entropy loss + mean squared error as its loss function. The generator and discriminator play a min-max game, and try to fool each other. We use two separate optimizers for the discriminator and generator. Here, the optimizers are based on Adam.
Use Cases:- Security, Accident Victim/Culprit Recognition, etc
Image Captioning refers to the process of generating textual description from an image – based on the objects and actions in the image In this project we have used the concept of Image processing and Natural Language Processing to develop an image captioning code. Image captioning is never 100% efficient because of the world and surroundings. In this project we have used a attention based model which enables us to see what parts of the image the model focuses on as it generates a caption. We have used the MS-COCO dataset ,preprocess and catches a subset of Kolkata/images using Inception V3.
This project requires the use of an accelerometer to constantly update the hand wearable orientation and simultaneously use a wireless communication system like Bluetooth to send the required trigger response to the connected machine. Our machine is also customisable viz. the user can customise the button simulations at any time using a pretty simple software provided free by us.
The key-points of our product are:
Inexpensive but can boost up gaming experiences.
It is completely a plug-and-play device. No need for an extra driver, software.
It is cross-platform. It works seamlessly with Windows, OSX and Linux.
The low learning curve to get adapted with the device.
The device fits well in hand and is very easy to use.
Easily customisable. User can change the controls at any time using our free software.
Low power requirements. The device uses Bluetooth Low Energy. It can give an active time of more than 50 hours.
The use of robots in the present day has moves from industries to the normal day to life. The use of voice commands to control a robot is much easier for domestic as well as industrial users. Voice controlled robotic system is also very beneficial in areas where there is high risk for humans to enter. This project i.e Arduino Voice Controlled Robot with Arms Implementation is controlled via Bluetooth module through a Smartphone app using Arduino. It is a step ahead of the robots of today and would definitely leave an impact on you.
humanoid robot is a robot with its body shape built to resemble the human body. The design may be for functional purposes, such as interacting with human tools and environments, for experimental purposes, such as the study of al locomotion, or for other purposes. Humanoid robots are now used as research tools in several scientific areas. Researchers study the human body structure and behavior (biomechanics) to build humanoid robots. On the other side, the attempt to simulate the human body leads to a better understanding of it. Besides the research, humanoid robots are being developed to perform human tasks like personal assistance, through which they should be able to assist the sick and elderly, and dirty or dangerous jobs. In essence, since they can use tools and operate equipment and vehicles designed for the human form, humanoids could theoretically perform any task a human being can, so long as they have the proper software. Elias is one such mini-humanoid that can interact with its surroundings.
Today we live an era of technology. With the rapid progress in the field of technology, today IoT, Data Analytics and Artificial Intelligence have ingrained themselves in our daily lives. The advent of these technologies in our lives also means that our vulnerability to various online and cyber threats has increased multifold. This makes security against cyber threats not just a luxury but a necessity in our lives today. While there are some tools in the market that serves as your defence against cyber threats, most of them are expensive and thus not within the reach of majority of people out there. This project, on the other hand, consists of a completely free and simple tool to detect malicious urls/websites by using a combination of free tools and then using an unique scoring algorithm to determine if the input url is safe or not and then let’s users visit the website. Though made using free tools this project is sure to provide you security that is at par with any expensive tool on the market.
Augmented reality (AR) is an interactive experience of a real-world environment where the objects that reside in the real-world are “augmented” by computer-generated perceptual information, sometimes across multiple sensory modalities, including visual, auditory, haptic, somatosensory, and olfactory. The overlaid sensory information can be constructive (i.e. additive to the natural environment) or destructive (i.e. masking of the natural environment) and is seamlessly interwoven with the physical world such that it is perceived as an immersive aspect of the real environment Augmented reality has been explored for many applications, from gaming and entertainment to medicine, education and business. Example application areas described below include Archaeology, Architecture, Commerce and Education. Some of the earliest cited examples include Augmented Reality used to support surgery by providing virtual overlays to guide medical practitioners to AR content for astronomy and welding. In this project we apply AR to further the works done in the field of Chemistry. It is an augmented reality based application that lets you combine elements using markers and view the valence electrons and the mechanism to form the compound.
Today, in India, there is a scarcity of commercial and housing space. The population explosion has ensured that the demand for space is perpetually more that the availability of the same. This largely contributes to the real estate values which in turn leaves a lot of people lacking proper housing. A flexible Living Space has dynamic flooring in which tiles can move vertically up to 2.67ft and adheres to users’ needs. The structure also has modular grid window where one can shift windows within the walls and change the shape and size. This ensures that this space can serve multiple purposes and thus do at least a little in curbing the rising problem of lack of space in the country.
With India being the second most populated country in the world, it’s only obvious that the traffic in this country would also warrant a mention of it’s own. India has the second largest network of roads in the world, yet traversing these roads is no less than waging a war for the citizens. One of the reason behind this is the rapid increase in the private ownership of vehicles and the overburdening of roads in all major cities. For an average Indian, the everyday struggle of dodging traffic and rash drivers has managed to become one of the major causes of chronic stress and high blood pressure. Thus, this project – a Deep Learning based approach to calculate traffic Congestion from camera feed, is aimed at making people aware of the traffic in a certain area and thus trying to prevent the inconvenience caused by it.
This project is composed of a car being controlled by hand gestures. Categories: Arduino, Electric Vehicles, Electronics, Engineering
A remote control trolley with versatile application like easy transporting, hotel services purpose. Easy to provide power with efficient outcomes. Categories: Electric Vehicles, Electronics, Fun & Games, Robotics
An AI that generates novel poetry when any word is tyed in the GUI. Categories: Engineering
Compact 4 wheeler vehicle made for twin riding with easy and safe riding conditions with 220cc capacity which can run at a speed more than 90kmph. The main attraction of the vehicle is it's look which is like a bike. Categories: Engineering, Electric Vehicles, Fabrication
It is a completely autonomous drone having great flight capability, ultra-stable flight, and can hover and perform different acrobatics in the air. Categories: Drones, 3D Printing, Flying & Aeronautics, Robotics
Formula type racing car designed and fabricated by the students. Categories: Engineering, Electronics, Fabrication
A tool which notifies you about any object it recognizes in its vicinity while you are away. Categories: Computers & Mobile Devices
This advanced project allows a user to control a fire fighter robot equipped with fire extinguisher remotely wirelessly for extinguishing fires. For this purposes the system uses an Rf remote for remote operation and wifi camera.. Categories: Robotics
A Bluetooth car. This car have two dc motors at its front and rear side. Front side motor is used for giving direction to car means turning left or right side (like real car steering feature). And rear side motor is used for driving the car in forward Categories: Arduino, Electronics