Get Ideas

The Snapdragon Neural Processing SDK gives developers the tools to accelerate their AI solutions on Snapdragon, and deliver intuitive and engaging deep learning experiences. Check out the entries below to explore creative ideas being submitted. We have also created Example Submissions to help you know what to include.

Submissions

Badriyatud Dini

Smart Pain Levels Recognition

Description

Assessment of pain levels based on facial mimic/expression changes has been done subjectively for the need of pain therapy. There is different results among the examiners because of subjectivity of examination. It needs a tool to analyze the pain level based on patient mimic for fast, accurate and

Ivo Georgiev

Consensus 3-D localization for mixed reality and autonomous

Description

Use Snapdragon NPEs to train distributed convolutional models, mapping multimodal sensory fusion (GPS, TOC, etc) of high-traffic spatial volumes to ultra-precise consensus volume localization, for use with high-fidelity mixed reality applications and reliability and safe autonomous robotics.

ahmad yulianto

Medical Records Biometric Identification

Description

The problem of searching patient data by name on Medical record system is duplication results. Proposed search by photo extracted from video mode which will in a real-time match up with the existing data. Using Neural Processing to reduce time in matching facial vector and get high accuracy results.

Yu-Jen Lin

Jpeg translators, since google is not doing it, we will

Description

jpeg, file that memes are made out of. No, this is not fully about that, I want to make a program that translate jpg. We have all this information stored in jgp/pdf put away, but never used, from books,memes, to websites, all we can do now is copy and paste it into translators.

Avinash Joshi

Intelligent phone assistant

Description

A phone assistant that can read out web articles from a browser when in her pocket and switch contexts to say reading out new mails/messages or playing music based on push notifications or voice commands from the user.

U Ukah

Private Tutor

Description

Develop an application that functions as a private tutor for students

Yogi Astawan

Smart Food Ingredients Identifier

Description

The system consists of 2 main components, there are the portable Spectrometer UV-Visible-IR and the Android application. Portable Spectrometer is the device used for identify ingredients of a food. The Spectrometer is using Ultraviolet-Visible Light-Infrared as light source and Photodiode as detect

ming huang

Self-learning surveillance system

Description

A surveillance system built on a Android phone, capable of recognizing family, friends and neighbors but detecting intrusions by strangers.

Sameer Kunde

On-device Federated Learning

Description

On-device ML is possible & necessary today due to increasing computing power on edge devices & burgeoning new use cases of ML. Federated Learning is an on-device ML architecture that can improve scene detection and language processing on edge devices while protecting user data privacy.

slim hmidi

cyst detector

Description

Develop an application to detect a cyst on on the mouth of a patient (under teeth).

Jewel James

Digients

Description

Digients are virtual pets on your phone that you can spend time with. You can spend your time with them talking or playing.You can feed/pat them when they do things you like and prod them away from things you don't. Over time, each digient learns/evolves to be the pet their trainer wants them to be.

Shreetam Behera

Automatic Scene Anomaly Detection

Description

Enhancing the capability of smartphones to detect anomalies in IP camera videos at home, public places, roads and crowded areas

Sumantra Dasgupta

The Enhanced Recommender

Description

Develop a personalized, content-based, collaborative, location based, AR enabled product recommender system capable of learning from user recommendation conversions and other audio/visual cues. The user can use audio,video, typing or written-note to input required item/product list.

Kishore Kanala

Personal Healthcare Assist

Description

Medical wearables including patches are getting very common. Data captured by these wearables and patches will be useful to draw inferences on health-fitness conditions and recommend medication and when the user need to consult doctor. Additionally it can suggest the diet such as food and water

bhargava reddy

using of A.I in hair styling

Description

an application which uses A.i to predict best suitable hair style for the user and show it in 3d

Manish Jaswal

Survivors' Scanner Toolkit for Immediate Response Team(IRT)

Description

When a disaster strikes , Immediate Response Teams usually are the first to reach the affected site, and their main aim is to minimize the possible loss of life.Due to chaos on the affect site, Manual scanning may delay the identification and can be fatal for possible survivors.

Robin Cooper

Emergency First Responder Automated Assistant

Description

To assist in the response to an emergency event I propose developing an app that will walk a first responder (FR) through the appropriate stages of patient assessment. The program would gather information from the FR and suggest treatments and record event details for future care.

Ashwini U

Language translator

Description

I want to build an offline real-time language translator which is useful for people who don't know foreign languages. This gives freedom to people who like to travel and explore new places. My app would be a voice based input-output system, which does real-time translation of audio input.

Pedro Olivera

Tutor for toddler

Description

I would like to build an App that can identify when the user learn a subject and then introduce a new subject. The app will target kids first learning subjects, such as colors, shapes, animals, numbers, letters and continue with more advance subjects as the kids master the subjects.

Juan Oosthuizen

Task Driven Companion

Description

The Task Driven Companion will analyze task specific mobile behavior and performed actions with the ability for added "snap recordings" (voice/notes/screen capture/results). Whenever similar task behavior is encountered in future the previous "snap recordings" can be used to enhance the outcome.

Abdulbasit Anate

CompanionBot

Description

CompanionBot is a Vision and Speech enabled Bot that is capable of seeing and communicating with a human like a consultant using machine vision and machine learning concepts.

Glenn Harmon

Speech Enhanced Multilingual Reading and writing

Description

Use google translation air buds to learn another language by displaying the translated language in the user native tongue on a smart phone.

Markus Bisselik

Bikeye

Description

Bikeye helps you to ride safer. It scans the road ahead, detects possible hazards and alerts you beforehand. Bikeye turns your existing mobile phone into a „Third Eye“. The AI powered app act as a portable collision avoidance system and road surface irregularities detector.

Payal Mulik

My Photo Gallery

Description

Develop an application to display a photos store in mobile based on voice. Photos should be displayed based on context provided by the user.

Anushtha Kalia

faceMed

Description

Early diagnosis of a disease remains a huge challenge in the healthcare sector. My motive is to build an application which can detect diseases with visually observable symptoms.

Sanjeev Dubey

SIEVE - Personalised Video Player app using SNE

Description

It is evident that what sort of content we consume controls our life. We must filter our content before consuming it just like we do with food. Today when web & social media videos creates chaos and emotional imbalance, SIEVE would bring a personalised and peaceful experience.

Utkarsh Mittal

Darwin - Intelligent Candidate Selection

Description

Darwin is an app that enables employers, organisations and institutions to rank multiple prospective candidates by conducting interviews using the app's platform. The platform analyses the interview video feed and EEG data from hardware (BCI) shipped with the app to rank the candidates.

Kobe Yu

Who Are You People

Description

In some public area need a real time face detection and recognition system, in the past those function are implment by Viola-Jones for face dection and recogntion by deep neural network. My idea is that combine two neural network into one, and using SNPE to speedup performance.

Juraj PANCIK

Grandp_AI_Ear

Description

The AI based system for monitoring the household of alone old people.We propose to create inexpensive electronic aids for social workers or family members.This low cost aids will long time monitor of life old alone people and draw attention to the deviations from the their normal days regime.

Vance Braziel

Deep Cyber Analytics Machine (DCAM)

Description

Welcome to the world of the DCAM. The deep learning capable cyber analytics machine in theory would be smart enough to realize its connected state and begin to learn about its network environment on its own. Eventually learning enough to establish a secure network free from vulnerabilities.

Scott Slack

Outdoor AR using signs and CV for SLAM

Description

Traditional AR has required AR Tags to determine position. Consider a OUTDOOR AR platform that with some knowledge of your location, can get an exact position, and run an AR application.

Himanshu Kumar

MoneyBook

Description

MoneyBook is a social trading App which lets user trade in stocks by converting daily and historical trade data into usable intelligence.

ruben jt johansen

Controller for institutional wheelchairs

Description

A microcontroller controlling the mobility of electrical wheelchairs within a predetermined area using GPS or WIFI signals in for example airports, hospitals, sports areanas and psych wards.

Emmanuel Asante

Personal Assistant - VIC

Description

What if you have a phone that is smart enough to keep a conversation with you. Not a voice recognition app or just a face recognition app but a combination of both. VIC can recognize emotions and voice of the user. Can also answer a call on your behalf.

Bill Richert

Diagnosing Patient Critical Care

Description

Use machine learning to diagnose patient critical care follow surgery, disease or injury.

Kaivalya Rawal

Predictive Calling for Phones

Description

Imagine a phone that can predict your commands before you give them. This project integrates personalized contact prediction with your phone's calling service. Each time you're making a call, you can select a suggested contact instead of typing a hard-to-remember name. Suggestions improve with time.

Christopher LA FONTAINE

BUSINESS TRIP ORGANIZER

Description

Bito will propose an organization according to meetings scheduled in your agenda.

Nanda Nandigavi

Know About Trees Around You

Description

The Idea is to have an app, which can scan the tree as a whole, or the leaf and give complete information about the tree. Which really helps the travelers ,Nature lovers and kids to know about Nature.

devesh j.

User centric smartphone

Description

Machine learning should open all the relevant apps based on user's usage history and put those apps on top of the display. On the basis of time and location, the ML on smartphone should help the user get most out of their phone without the user input.

Nitin Bhaskar

Stock predictor

Description

Idea is predict stock prices in near future of listed companies so that the profit can be maximized. This is possible with machine learning. Prediction is mainly based on past price variation trends and other affecting factors which can fetched using news feeds.

LOGESH WARAN

Supports with Medical Instruments and Camara Technologies

Description

For the Purpose of medical instruments for quick and compact-able access with no hazards to the patient... and the Camera uses to like the realistic scanning abilities with the wireless model. And then the halo-graphical display... like 2D

Priyadarshan Patel

Self Driving Car

Description

Self driving car is the hottest topic in Silicon Valley. MIT students with collaboration of Nvidia Jetson's boards developed RC cars which does racing. Similar idea is to implement and run the car with all the capabilities of Computer Vision and Sensor fusion.

Pierre-Emmanuel FEGA

Madagascar Transportation Network Live Improvement

Description

My project is about improving Madagascar Transportation Network. An AI-enabled Snapdragon Neural Processing Engine device will analyze the condition of the malagasy roads. Image and lidar feeds will be submitted to an ML model inorder to help authorities to take proper roads improvement decisions.

Vincent Marius

Guessed Video Frames

Description

For devices with bad internet connexion, we could create a video format with less images than in the original, and the AI would be trained to guess the missing ones. The video could be converted with a neural network, and read from the client.

kishore gadhi

Clearing of RAM and Cool of Unit

Description

with my concept i want to create a DL algorithm based on which the unit will clear the RAM space and cool the unit.

vasu kokkiligadda

Auto Response generator

Description

This will help to generate and auto suggest response based on the conversation going on currently in any chatting application, Learning from the user's past response style and enhanced with cognitive sense. Helps user's in having a quick conversation.

Sunil Kumar

Educational application

Description

I want to make application that rum on android operating system, which provide the educational data of all subjects with the workplace for school students. This application help to saving paper for student and parents can track the student knowledge criteria.

Laszlo Dudas

j0shu4

Description

Ai friend with dreams.

Barathwaj C

The human

Description

The AI to be your best friend. calm and console if you are sad, celebrate and make fun if you are happy or joyful. cool and relax you if you are angry and frustrated. share and care the fundamental human emotions with you and be your best mate at all times.

Mounika Gude

USER BOT

Description

Using NLP technique I want to make an automatic call at the time I have fixed.On mobile the bot itself has to see the time and date and have to call the selected contact and speak to them without human intervention. ans should be able to briefly summarize the call.

Ajay Kumar Jaiswal

Rash Drive Identification (Stunts ,Skidding,Speed)

Description

To restrict people from stunt and rash drive. The ML Algorithm can compare with model and tell the status of driving (Stunts, Front wheel drive, back wheel drive,skiding).

Sagar Pomal

It's now easy to remember faces and names.

Description

The phone/glasses with help you to recognize people.

Taruna Agrawal

Selective printing of objects from mobile

Description

Selective printing from mobile enables user to print selective objects (texts, images, tables etc.) from documents without the need to select and crop manually from the document.

Thejaswi pallapothu

Addiction Checker/Preventer

Description

A lot of people are addicted to lot of things and most of them wanted to change. Machine Learning can give them chances to change.

Yogaraj Rajendran

All I Need To Know

Description

All I need to know what may happen nearby. An app(AI) that learn about user(people, machine etc), talk with nearby AI using IoT and prognosticate accordingly.

Ajay Prasad

Carzy Bot

Description

The future phone should need an AI.But the future phone still need a bixiy button or Siri. This is definitely not Make a future phone.Take off all the stupid button and do with gesture.The Gesture make the future phone.And it give a new shape to phone.This is what I thought!

Fahad Kabli

Sensual perception compute engine

Description

To intersect with modern environment sensual technology is needed this will change the way of interaction with our modern devices.

Ayush Mittal

Indoor Localization

Description

I am building a mobile application to be used for indoor navigation. The application does not require any special infrastructure installed in buildings. It will use Wi-Fi signal fingerprints to predict location in buildings using Machine Learning models and support real-time navigation.

Kiran Srinivas Rao

Bookz

Description

An app to extract Title, Author, Publication, Date and version from physical books, maintain a database - to store, update and remove it online- make friends, family and like minded people to share, buy or sell physical books online. Sort it and arrange it using external hardware.

Ruchit Nigam

AI Guide

Description

People in many villages and small cities as well as few people in cities don't know much about how to operate their latest smartphones. So there could be a AI playform that can understand many languages so that people can be guided on how to work in their smartphones smartly and effectively

Ashok Nallamilli

Personal Assist using AI & IoT

Description

Providing assistance by audio/video/control signals(IoT) to user by monitoring user usage in mobile,laptop,fitness trackers and other smart systems.

Naveen Kumar Valluri

Understanding the daily life of the particular user

Description

Understanding the daily life of the particular user and should able to manage their daily needs

Nyron Morrison

Pro-T31n

Description

There are a number of AI Apps/Devices already in development, created to make everyday tasks simpler and more interactive but there are still many different tools to be created. My idea revolves the human body itself rather than the mind.

Yash Raj

Sense behaviour make mobile sensor work

Description

Sense behaviour of our body will able to make the smart phone sensor work using AI like we can use in Google Assistant .we can able to unlock or use the sensor like moving hand to unlock our phone or eye sensor to open social chat ,etc

Sarbjit Hanjra

Canny Buddy

Description

I am looking to develop a smart doll for kids. I am having two beautiful daughters and I found that they like to spend time with their dolls. They even talk to their doll, they play with them. I am willing to create a buddy for them so they can learn and teach from smart buddy.

Kyle Birchard

Automated ID of airborne pests and disease vectors

Description

This application allows autonomous aerial vehicles to identify, pursue, and eradicate airborne insect species posing risks to public health and crop production. On-board AI allows species identification, as well as making changes to flight path based on complex geographic and real-time weather data.

James Reid

Keepup

Description

This is an application to help with daily tasks. It is intended to target users that may have memory impairment or difficulty scheduling important daily tasks such as maintaining a medication regimen. Additional capabilities include safety monitoring and authorized user remote access.

Andrew Enright

Utilize Built-in GPU

Description

Repurpose the graphics processing on devices that already have it.

Rutvik Panchal

Game Learning and its Implementation (Self Driving Cars)

Description

In Context, First of all i will to design a program that will learn how to drive a car in realistic open world games such as GTA V and will store that info. in a database and then i will feed that data to SoC featuring Snapdragon enabled device and that will be able to drive a car in real world.

Aneev Reddy Y

Busy makes call perfect

Description

Know before whether the caller is busy in meetings based on his calendar. This should be helpful when multiple callers are calling when someone or busy.

sheshank arul

bionic navigator

Description

An bionic arm that will sense the moments and visuals and that can be used from far distance from where it is operated.

Jacob Small

Ultimate Fact Checker

Description

By leveraging "The Wisdom of the Crowd" Wikipedia is able to distill information down to a semblance of fact. We live in very uncertain times and with the prevalence of "Fake News" the substance of truth is questioned daily. Let's take crowd sourced wisdom and use it to teach an AI to find truth.

slim hmidi

psychoPlace

Description

Propose for someone a best place to visit related to his emotion and psychological situation.

Nitin Bhaskar

Navigation for visually impaired

Description

Visually impaired people find it very difficult to navigate around places. Relying on assistance (people who can help) with directions is not an optimal solution, as there might not be people around always. Sometimes walking too becomes very difficult and painful as they might miss stones/potholes.

Elavarasan Ramasamy

AEYE

Description

This project will use the image processing algorithm and will guide the visually challenged people in the way even any one can close their eyes and walk by using this system.

Sergei Popkov

The House with Conscience

Description

I would like to make special kind of "Clever house" - House with Conscience, controlled via phone, which can develop personal feelings related to its owner. The House can 3D-print the gifts to his owner or try to "kill" him - stop the services and allow the owner to meditate about his spoiled ways.

Aishwarya Das

Netrayan

Description

Netrayan app for street-guidance uses visuals from Street View which lets users search for objects on their Google Maps route. For instance, a user can look for benches or parks along their route. The app can also guide visually-impaired users with road conditions such as obstacles, road signs, etc.

Clayton Morse

The smart assistant.

Description

The smart assistant will not only be able sort out your daily planes but also be able to communicate with your smart appliances and other micro controllers.

rahul kukreja

Artificial Intelligence

Description

we need to develop such kind of AI for pen who can answer question while writing in exams. If we can develop such concept there will be more chances to make a pen who can write on its own , by just giving instruction we would make a new era of technology and development.

Manbir Jhawer

LeNet neural network

Description

I would implement LeNet neural network on the qualcomm ai chipset.

Simon Mayrhofer

Fooderingo - The Intelligent Ordering System for Restaurants

Description

I would create a smart Ordering System for Restaurants, for example a Easy-Order that is smart and knows what the clients want to order.

Pinto Felix

Ocean navigator

Description

To empower the visual search capability of search vessel for endangered marine life. Various species under water are effected due to human acts. Now they are searched and tagged for protection. The task is to visually identify, tag and classify marine mammal by 360 degree images underwater.

sai reddy

Wake locks

Description

My idea is to increase battery life which we call efficiency of a Soc. Recently to increase efficiency Qualcomm introduced Big.little technology. But my idea is related to application wake locks. By preventing the wake locks of applications that user don't use frequently. We can save battery life.

Angel Barranco

Automatic app organizer

Description

My idea is creating an app to improve usability of Android apps by automatically organizes the apps on home screen. By time of day and patterns of use.

Tom Brereton

Handicap Helper Bot

Description

Autonomous Handicap\disability robot with the ability to follow the person indoors/outdoors with minimal input to assist in supplying oxygen, monitor vitals, call for help if needed and log health data.

Jerry Pettersson

To be friend with the accountant

Description

A camera function, that recognize receipt and invoices credit-card slips airplane tickets, understanding what is receipt is for. Sorting them in order and in categories. Helping you be a star, regarding your travel expenses report. Your shopping habits ....... and creating a you wonderful report.

Kaveh Jokar Deris

Face recognition security cam

Description

The idea behind this project is to build a surveillance system which uses AI and face recognition models. Once AI is trained with familiar day to day faces, the cameras will only target and follow the unfamiliar faces.

Stephen Lee

Notework

Description

All note apps fall short in terms of automatic organization, search, and convenience. Notework is a new kind of note taking app. It offers automatic labelling, advanced search, layout, and merging of notes. It allows you to collaborate conveniently, and tag URLs. It does OCR on your notebooks.

Gautam Saha

USTAD

Description

USTAD (User Service for Training Advancement and Development) is an immersive training platform designed to provide guided instructions as the user interacts with the system via a suite of domain specific applications. The host 835 system will obtain input from connected devices and guide the user.

Louis Vichy

Smart Decentralized Cache

Description

Using data gathered from users in the network (with permission), the model predicts future data a certain group of user are likely to need from remote stores. A program then use these prediction to cache or release the data appropriately based on current user state.

Sri Harsha Seethalam

Explicit Content Detection - Parental Control Application

Description

Parental Control Application - Explicit Content Detection on Android Platform. Uses deep learning for image classification of explicit content to safe guard kids.

AJIT DESHPANDE

Pocket Doctor

Description

Intensive applications utilizing Machine Learning for medical diagnostics can be executed locally on mobile devices. These ML on-device apps clubbed with various sensors on a mobile device can help classify different medical conditions,right after a measurement done with on-device sensors.

Onofrio Panzarino

The Risk Expert

Description

The Risk Expert classifies how much safe is a place to suggest behaviors and protection equipment. It is based on historical data, place geometry and can detect new risks.

Dijon Vula

SensorSenseAI

Description

The app (or lower level application) would analyze datasets and logs of the smartphone sensors and make predictions of the usage by the user. The predictions would load apps in RAM before use, clear some apps, do specific tasks all by predicting user behavior by the sensor data trained AI.

Amit Saraf

Intelligence In Agriculture

Description

The goal is to replace the traditional way of decision making in farming. We can use Drones surveillance to capture the images for soil fertility, existing irrigation, crop disease. The cognitive intelligent system will analyse capture information to generate accurate report suggesting action plan.

Melvin Chotu

Awake Driver

Description

The CDC found that 1 in 25 adults have fallen asleep at the wheel. I believe that Artificial Intelligence could end this. With facial recognition technology, the ques of a sleepy driver could be detected.

Madhan Raj

Check water cuurent

Description

To check the water current in a particular spot of the ground to dig for bore well to draw water. This will be very helpful in drought areas where water is a scarcity. AI will help to develop a method of accurate measurement.

Spencer Leamy

Eagle3ye

Description

Eagle3ye is a rapid prototyping platform for GPU accelerated 2D/3D computer vision algorithms. It has an easy to use GUI designed for fast and flexible model driven development that saves processing pipeline configurations into a portable container for quick assembly of algorithm components.

John Wright

Smart Garden

Description

The Smart Garden will monitor the nutritional needs of plants and environment in a hydroponic greenhouse to provide the optimal nutrients real time. It will also monitor plants via camera and notify users when their plants need extra attention from pests or disease before its to late.

Shubham Gupta

Control over system

Description

To develop an AI project which can first recognize the voice of the authorised person, then we use speech recognition for access the command given by the user. This speech recognition will help to have a control over our computer system, without using of keyboard and mouse. To access the systems.

Isaac Osei

Remote Elderly Healthcare with IOT

Description

The goal is to remove the barrier to using sophisticated technology by elderly and seniors using remote operations by intelligent systems and highly trained personnel

Damian Drygalski

IBD Triggers Finder

Description

Discovering inflammation triggers for IBD patients.

chris dryden

Don't I know you?

Description

This application would process a video feed in real time and apply an augmented-reality style overlay that highlights individuals and provides information about who they are. Incorporated into a Google Glass-style wearable device this could help you remember where you saw that guy before.

Latha Pushpanathan Arun

Virtually choosing Paint for Buildings

Description

This is to visualize building interior and exterior by selecting colors from shade cards and to give accurate image of after-paint home. Also to measure height, width of a wall and estimate quantity needed to paint.

Deeptesha Satpute

Prediction of student alcohol consumption app

Description

The app will predict whether about the alcohol consumption tendency of the students based on the information that student provides.

Craig Campbell

Applied Behavioral Analysis Autism Therapy Application

Description

Imagine an app that has unlimited patience making autistic/normal kids sing and play in ways they think is fun but they are actually learning language, ABC's and math. An app a therapist can give a kid that has unlimited patience and requires specific input (speech and eyeball detection) to proceed.

Patricio Martinez

Professional Assesment for your Properties

Description

Web and Mobile application for registration and assesment of private and public properties. The information retrieved will be used to generate a commercial and sales value for owners and banking systems. It requires comparison with nearby properties, facilities and specific features of the property.

Rakshith Shanbhag

Robotic Umpiring System

Description

The idea is to eliminate manual umpiring system in any sports, since there are lot of false decisions made during the course of action. The robots can be trained by adapting techniques of machine learning and image processing. More accurate decisions could be made instantly at real time

prathyusha mallela

musical note writer

Description

The idea is to write out music in different notations (example staff notation) for musicians/amateur instrument players for a given song.

Frank Salvo

The Companion

Description

Your mobile machine will constantly monitor you so that it guides your day. Reminds you of everything, from A to Z, in that day and every day that you ask it to be with you. Always in a constant dialogue as if it is your better self with all of your habits. Like talking to yourself.

Carlos Olson

Amicus

Description

Wouldn't it be awesome to have your own personal friend that just knows how to ask you that right question. Amicus is that friend.

Tzach Kaufmann

Future Explorer

Description

Future explorer will help you make better decisions for your daily life

Brijesh Sharma

Proactive Behavior Management

Description

This will Learn from action of a person on phone or otherwise and will prompt the related near proximity users about the behavior which is expected out of this person.This will be usefull for Diplomats /Negotiaters /Management /HR /Boss / Husbands.

Eddy Valverde Garro

Resource Spending Measurer

Description

The idea consists of a network of devices recording how much water and how much energy we are using constantly, and how much we should be using. Saving our planet from collapsing is the priority.

Harishanth Raveendren

Structural prediction using IR

Description

Analysing the structure of a building requires more man power to understand the details. Using IR mounted devices we can calculate the dimensions of the building and effectively find if there are occupants which can aid the military trying to explore unused building for terrorist in habitation.

Arun BV

Robotic Home Paint System

Description

The brief idea of the proposal is that making the job of painting house simple. The robots are trained to paint the wall by specifying the area of the wall and specific color combinations. The arm movement of the robot must be calculated by using some simple mathematics to get the desired result.

Abhijna Kashyap

Object Measurement

Description

The application will be able to detect any object Length, Width, Height from mobile camera. This AI app will be able to detect different objects around a person, For example to know the measurement of a table,box etc just hold the app camera towards it and measurement will be displayed in screen.

Alexandru Pruncu

AI Explorer

Description

The application will use the power of a smartphone to transform a robotic platform into an intelligent agent capable of pursuing predefined goals in an environment.

Brian Eshelman

Microclimate Weather Predictor

Description

Use common available weather station inputs in a residential package to predict weather for a very specific location.

Prince Patel

On device surveillance - Smart City

Description

Millions of CCTV camera's presently are just recording devices over cities. With efficient inference of neural networks on device, the idea is to empower surveillance on device(CCTV) without being connected to cloud for computer vision tasks.Making artificially intelligent camera is the goal.

MURALEEDHAR P

GoodMark

Description

An app that bookmarks everything. Imagine you are travelling and see some useful information. You click a pic. The AI in your processor will be knowing that suddenly you have started the camera and took a pic. The details of the pic might be useful later like location of the pic,etc.

Bofu Chen

BerryNet on Mobile

Description

BerryNet is a FLOSS project turning Raspberry Pi 3 into an intelligent edge application with deep learning running on it. BerryNet on Mobile project will bring more possibilities to democratize AI in our life.

Manu Konchady

An Automated Language Trainer

Description

This app will be used to learn English, however the main ideas of the app can be applied to other languages. Machine learning will be used to analyse a given tagged corpus to generate a large number of rules to detect punctuation. Then quizzes will be generated from uploaded text.

CHALAVADI GOWTHAM SAI

Interconnecting devices

Description

Interconnecting devices means an interconnecting of devices in according to human needs, depending on the profile he build and the past activity , recently visited like movies, coffee shops and all with respect to the information source available like showing offers ,playing songs and so on.

Bill Hicks

ML CC Reader

Description

Current credit card readers that use the phone camera (such as Card IO) use opencv for graphics manipulation which adds 15 to 20 megabytes of library code to the apk. A ML algorithm should be able to recognize and extract the card information faster using 1/10th the size.

Venkatesha Manja

Intelligent assistant which learns

Description

Intelligent assistant app provides timely suggestions/notifications using various inputs.App also learns from usage patterns/past behaviour and updates weights for future suggestions/notifications accordingly. Ex:Suggest to follow up on an email/life goal while sitting in a train/bus, waiting/jog.

Jenish Baldha

Predicting handwriting pattern

Description

based on image of sample words, predicting handwriting for every Unicode character

N Vijay Anand

Safe driving

Description

This app will help detect the performance of driver and raise alerts for negligence driving and hence aid in minimising traffic accidents.

Henry Moseti

Personal Critical Situation App

Description

This application will determine heart rate and sounds to identify critical situations. For example, there may be robbery or terror even domestic fight. The app would recognize the difference between the data and create different alerts. It would even recommend the next based action or automate it.

Rohit Bhaskar

ClearPath - Utilize travel to the fullest

Description

Traffic is a huge problem all across the world, especially in India. It is estimated that the average Indian loses almost 2hrs a day in travel. This is almost 1/10th of your day. Our app will understand your job and timings and make use of machine learning and games to make you more productive.

Deepak Sood`

Pick & tick the Road

Description

Roads- major part of countries economy are constantly used which results in deterioration of the roads and cracks come on it. there is crack detection and presentation systems exist but are way to expense and are not enough.The Implementation that is conceptually till now will be used.

Prasanna Mondkar

ViFil - Video Content Filter

Description

Video Content Pattern Matching App, which looks for nudity (porn), violence (blood, dearranged body parts, dead bodies), disturbing content (screams, abuses, bad words, etc.) in a video and provide feedback regarding contents (eg., violence - 40%, Nudeness - 80%, abusive language - 90%, etc.)

Tejas Dhagawkar

Cartoonist

Description

This application will take photo of the person and then create 3D mesh of it using neural network. These meshes then can be used in a games or user can apply predefined skeleton animation to them. The character will resemble to the person in the photo. User can export the mesh and share it.

Abhilash Laguri

Gandhari

Description

Use of DRONE to guide a visually impaired to navigate through his journey, your phone will work as a mediator between the drone and the user.

Peter Krall

Pipeline leak detection

Description

Small robot submarines will inspect ageing water supply systems. They will simultaneously measure fluctuations of flow and sound, both caused by variation in demand. Superposition of the patterns will allow mapping the supply system and identifying the sections needing repair.

Tarun Yellogi

EB Intelligence

Description

Behavioural Intelligence - plays a important role in finding,observing,categorisiging ones actions. Emotional intelligence - is important to identify ones feelings in any situation, usefull to identify their behaviour based on different situations. :Emotions will be in head - behaviour in action.

Mohamed Hassan

Nutshell

Description

To save user time this application give user a summary for what happen on the news, tweets, Facebook posts based on user interests. This will save user time and at the same time keep him updated

Mohamed Hassan

AI Sewrch

Description

Smart search for anything in all files types on the mobile phone. Result can be text files, images and clips. You don't need to look to all your images to find images that your child appear in it, all images contains river for example or clips that contains birthdate party seen

Pradeep Jain

OneBody4Life Smart Apps for Chronic Disease Management

Description

Our Android Apps (HIPAA/serverless) control BP, heart attacks strokes, control T2 Diabetes and COPD/Asthma - killing 60% of NCD globally. Can run on $35 phones next 2 billion free with 3 levels of Coaching, Speech, 63 Rules, predictive analytics, reminders. ML/NN engines is a premium feature.

Samuel Cruz

Writing Assistant

Description

Days ago during a messaging with a friend, we talked about about a meeting on a popular place but during the typing, I realize that would be useful that my keyboard after making predictions about the names and words also would be nice provide me stickers recommendations about text written.

Muni prasad K

Paddy diesase detection using image processing technique

Description

Initially Capture the paddy disease affected image.after its going to image processing to reduce the noise and to get the threshold values. Based on threshold values the images can be compared and classified with the existed images stored in Database. using Neural networks.and provides solutions...

Jose Johnen Gonzalez

Travel Optimizer

Description

the app check the places than you travel frequently on a map for some time and then purpose new alternative routes to do your every day faster.

Jeong Hwan Kim

Emotional Intelligence

Description

My phone understands how I feel, and how my friends feel and give advice for better relationship or self-awareness.

Need inspiration to help with brainstorming?

See the links below to some informative content.

Snapdragon and AI at the Edge Webinar

Gary Brotman, Director, AI Product Management, discusses what Qualcomm Technologies is doing in AI, and why we believe mobile is the perfect platform to drive AI experiences on-device.

Learn More

Introduction to AI and Snapdragon NPE SDK

Looking at AI through the eyes of a developer and the balance of performance and power consumption.

Learn More

Developing for AI on the edge

Exploring some example AI applications where on-device computation can provide differentiation and critical advantage.

Learn More

Resources

Get Inspired. To help you gather all the insight, information and direction you need, we have provided links to the following resources:

Qualcomm Developer Network is a community of coders, makers and thought-leaders. We’re working together to share ideas and inspiration with YOU and the next generation of high-performance experiences and designs.

Learn more

TensorFlow is an an open-source machine learning platform for everyone. It's a fast, flexible and production-ready framework that scales from research to production.

Learn more

Caffe is a deep learning framework made with expression, speed, and modularity in mind. Take a look.

Learn more

Caffe2 provides an easy way for you to experiment with deep learning. You can bring your creations to scale using the power of GPUs in the cloud or to the masses on mobile with Caffe2’s cross-platform libraries.

Learn more

About the AI Developer Contest

To enter the AI Developer Contest, describe in English and in 600 characters or less, what YOU would create with on-device mobile machine learning and artificial intelligence on a Snapdragon mobile platform. An optional photo may be included.

You have until October 13, 2017 to submit your entry. Throughout the contest all eligible entries will be posted on the webpage (developforai.com/submissions) for others to see. Entries will be judged based on creativity 25%, best potential use of the Snapdragon processing power 25%, and real-world application 50%.

Twenty (20) winners will be announced on October 20, 2017 and each will receive the prize of a Snapdragon 835 Mobile Hardware Development Kit. These 20 winners are eligible to compete for the grand prize trip for two to Mobile World Congress in Barcelona, by submitting an APK of their demo along with a supporting video bringing their idea to life, by no later than December 8, 2017.

Contest rules