IfcPlusPlus is an open source C++ class model, as well as a reader and writer for IFC files in STEP format. Features: Easy and efficient memory management using smart pointers. Parallel reader for very fast parsing on multi-core CPU's. Additionally, there's a simple IFC viewer application, using Qt and OpenSceneGraph. It can be used as starting point for all kinds of applications around the open building model standard IFC.

Package manager for things. Start designing modular hardware! PartCAD is the standard for documenting manufacturable physical products (a.k.a. Digital Thread or TDP). It comes with a set of tools to maintain product information and to facilitate efficient and effective workflows at all product lifecycle phases, boosted by AI.

using Neural Networks (SSD) on Tensorflow. This repo documents steps and scripts used to train a hand detector using Tensorflow (Object Detection API). As with any DNN based task, the most expensive (and riskiest) part of the process has to do with finding or creating the right (annotated) dataset. I was interested mainly in detecting hands on a table (egocentric view point). I experimented first with the [Oxford Hands Dataset](http://www.robots.ox.ac.uk/~vgg/data/hands/) (the results were not good). I then tried the [Egohands Dataset](http://vision.soic.indiana.edu/projects/egohands/) which was a much better fit to my requirements. The goal of this repo/post is to demonstrate how neural networks can be applied to the (hard) problem of tracking hands (egocentric and other views). Better still, provide code that can be adapted to other uses cases. If you use this tutorial or models in your research or project, please cite [this](#citing-this-tutorial). Here is the detector in action. <img src="images/hand1.gif" width="33.3%"><img src="images/hand2.gif" width="33.3%"><img src="images/hand3.gif" width="33.3%"> Realtime detection on video stream from a webcam . <img src="images/chess1.gif" width="33.3%"><img src="images/chess2.gif" width="33.3%"><img src="images/chess3.gif" width="33.3%"> Detection on a Youtube video. Both examples above were run on a macbook pro **CPU** (i7, 2.5GHz, 16GB). Some fps numbers are: | FPS | Image Size | Device| Comments| | ------------- | ------------- | ------------- | ------------- | | 21 | 320 * 240 | Macbook pro (i7, 2.5GHz, 16GB) | Run without visualizing results| | 16 | 320 * 240 | Macbook pro (i7, 2.5GHz, 16GB) | Run while visualizing results (image above) | | 11 | 640 * 480 | Macbook pro (i7, 2.5GHz, 16GB) | Run while visualizing results (image above) | > Note: The code in this repo is written and tested with Tensorflow `1.4.0-rc0`. Using a different version may result in [some errors](https://github.com/tensorflow/models/issues/1581). You may need to [generate your own frozen model](https://pythonprogramming.net/testing-custom-object-detector-tensorflow-object-detection-api-tutorial/?completed=/training-custom-objects-tensorflow-object-detection-api-tutorial/) graph using the [model checkpoints](model-checkpoint) in the repo to fit your TF version. **Content of this document** - Motivation - Why Track/Detect hands with Neural Networks - Data preparation and network training in Tensorflow (Dataset, Import, Training) - Training the hand detection Model - Using the Detector to Detect/Track hands - Thoughts on Optimizations. > P.S if you are using or have used the models provided here, feel free to reach out on twitter ([@vykthur](https://twitter.com/vykthur)) and share your work! ## Motivation - Why Track/Detect hands with Neural Networks? There are several existing approaches to tracking hands in the computer vision domain. Incidentally, many of these approaches are rule based (e.g extracting background based on texture and boundary features, distinguishing between hands and background using color histograms and HOG classifiers,) making them not very robust. For example, these algorithms might get confused if the background is unusual or in situations where sharp changes in lighting conditions cause sharp changes in skin color or the tracked object becomes occluded.(see [here for a review](https://www.cse.unr.edu/~bebis/handposerev.pdf) paper on hand pose estimation from the HCI perspective) With sufficiently large datasets, neural networks provide opportunity to train models that perform well and address challenges of existing object tracking/detection algorithms - varied/poor lighting, noisy environments, diverse viewpoints and even occlusion. The main drawbacks to usage for real-time tracking/detection is that they can be complex, are relatively slow compared to tracking-only algorithms and it can be quite expensive to assemble a good dataset. But things are changing with advances in fast neural networks. Furthermore, this entire area of work has been made more approachable by deep learning frameworks (such as the tensorflow object detection api) that simplify the process of training a model for custom object detection. More importantly, the advent of fast neural network models like ssd, faster r-cnn, rfcn (see [here](https://github.com/tensorflow/models/blob/master/research/object_detection/g3doc/detection_model_zoo.md#coco-trained-models-coco-models) ) etc make neural networks an attractive candidate for real-time detection (and tracking) applications. Hopefully, this repo demonstrates this. > If you are not interested in the process of training the detector, you can skip straight to applying the [pretrained model I provide in detecting hands](#detecting-hands). Training a model is a multi-stage process (assembling dataset, cleaning, splitting into training/test partitions and generating an inference graph). While I lightly touch on the details of these parts, there are a few other tutorials cover training a custom object detector using the tensorflow object detection api in more detail[ see [here](https://pythonprogramming.net/training-custom-objects-tensorflow-object-detection-api-tutorial/) and [here](https://towardsdatascience.com/how-to-train-your-own-object-detector-with-tensorflows-object-detector-api-bec72ecfe1d9) ]. I recommend you walk through those if interested in training a custom object detector from scratch. ## Data preparation and network training in Tensorflow (Dataset, Import, Training) **The Egohands Dataset** The hand detector model is built using data from the [Egohands Dataset](http://vision.soic.indiana.edu/projects/egohands/) dataset. This dataset works well for several reasons. It contains high quality, pixel level annotations (>15000 ground truth labels) where hands are located across 4800 images. All images are captured from an egocentric view (Google glass) across 48 different environments (indoor, outdoor) and activities (playing cards, chess, jenga, solving puzzles etc). <img src="images/egohandstrain.jpg" width="100%"> If you will be using the Egohands dataset, you can cite them as follows: > Bambach, Sven, et al. "Lending a hand: Detecting hands and recognizing activities in complex egocentric interactions." Proceedings of the IEEE International Conference on Computer Vision. 2015. The Egohands dataset (zip file with labelled data) contains 48 folders of locations where video data was collected (100 images per folder). ``` -- LOCATION_X -- frame_1.jpg -- frame_2.jpg ... -- frame_100.jpg -- polygons.mat // contains annotations for all 100 images in current folder -- LOCATION_Y -- frame_1.jpg -- frame_2.jpg ... -- frame_100.jpg -- polygons.mat // contains annotations for all 100 images in current folder ``` **Converting data to Tensorflow Format** Some initial work needs to be done to the Egohands dataset to transform it into the format (`tfrecord`) which Tensorflow needs to train a model. This repo contains `egohands_dataset_clean.py` a script that will help you generate these csv files. - Downloads the egohands datasets - Renames all files to include their directory names to ensure each filename is unique - Splits the dataset into train (80%), test (10%) and eval (10%) folders. - Reads in `polygons.mat` for each folder, generates bounding boxes and visualizes them to ensure correctness (see image above). - Once the script is done running, you should have an images folder containing three folders - train, test and eval. Each of these folders should also contain a csv label document each - `train_labels.csv`, `test_labels.csv` that can be used to generate `tfrecords` Note: While the egohands dataset provides four separate labels for hands (own left, own right, other left, and other right), for my purpose, I am only interested in the general `hand` class and label all training data as `hand`. You can modify the data prep script to generate `tfrecords` that support 4 labels. Next: convert your dataset + csv files to tfrecords. A helpful guide on this can be found [here](https://pythonprogramming.net/creating-tfrecord-files-tensorflow-object-detection-api-tutorial/).For each folder, you should be able to generate `train.record`, `test.record` required in the training process. ## Training the hand detection Model Now that the dataset has been assembled (and your tfrecords), the next task is to train a model based on this. With neural networks, it is possible to use a process called [transfer learning](https://www.tensorflow.org/tutorials/image_retraining) to shorten the amount of time needed to train the entire model. This means we can take an existing model (that has been trained well on a related domain (here image classification) and retrain its final layer(s) to detect hands for us. Sweet!. Given that neural networks sometimes have thousands or millions of parameters that can take weeks or months to train, transfer learning helps shorten training time to possibly hours. Tensorflow does offer a few models (in the tensorflow [model zoo](https://github.com/tensorflow/models/blob/master/research/object_detection/g3doc/detection_model_zoo.md#coco-trained-models-coco-models)) and I chose to use the `ssd_mobilenet_v1_coco` model as my start point given it is currently (one of) the fastest models (read the SSD research [paper here](https://arxiv.org/pdf/1512.02325.pdf)). The training process can be done locally on your CPU machine which may take a while or better on a (cloud) GPU machine (which is what I did). For reference, training on my macbook pro (tensorflow compiled from source to take advantage of the mac's cpu architecture) the maximum speed I got was 5 seconds per step as opposed to the ~0.5 seconds per step I got with a GPU. For reference it would take about 12 days to run 200k steps on my mac (i7, 2.5GHz, 16GB) compared to ~5hrs on a GPU. > **Training on your own images**: Please use the [guide provided by Harrison from pythonprogramming](https://pythonprogramming.net/training-custom-objects-tensorflow-object-detection-api-tutorial/) on how to generate tfrecords given your label csv files and your images. The guide also covers how to start the training process if training locally. [see [here] (https://pythonprogramming.net/training-custom-objects-tensorflow-object-detection-api-tutorial/)]. If training in the cloud using a service like GCP, see the [guide here](https://github.com/tensorflow/models/blob/master/research/object_detection/g3doc/running_on_cloud.md). As the training process progresses, the expectation is that total loss (errors) gets reduced to its possible minimum (about a value of 1 or thereabout). By observing the tensorboard graphs for total loss(see image below), it should be possible to get an idea of when the training process is complete (total loss does not decrease with further iterations/steps). I ran my training job for 200k steps (took about 5 hours) and stopped at a total Loss (errors) value of 2.575.(In retrospect, I could have stopped the training at about 50k steps and gotten a similar total loss value). With tensorflow, you can also run an evaluation concurrently that assesses your model to see how well it performs on the test data. A commonly used metric for performance is mean average precision (mAP) which is single number used to summarize the area under the precision-recall curve. mAP is a measure of how well the model generates a bounding box that has at least a 50% overlap with the ground truth bounding box in our test dataset. For the hand detector trained here, the mAP value was **0.9686@0.5IOU**. mAP values range from 0-1, the higher the better. <img src="images/accuracy.jpg" width="100%"> Once training is completed, the trained inference graph (`frozen_inference_graph.pb`) is then exported (see the earlier referenced guides for how to do this) and saved in the `hand_inference_graph` folder. Now its time to do some interesting detection. ## Using the Detector to Detect/Track hands If you have not done this yet, please following the guide on installing [Tensorflow and the Tensorflow object detection api](https://github.com/tensorflow/models/blob/master/research/object_detection/g3doc/installation.md). This will walk you through setting up the tensorflow framework, cloning the tensorflow github repo and a guide on - Load the `frozen_inference_graph.pb` trained on the hands dataset as well as the corresponding label map. In this repo, this is done in the `utils/detector_utils.py` script by the `load_inference_graph` method. ```python detection_graph = tf.Graph() with detection_graph.as_default(): od_graph_def = tf.GraphDef() with tf.gfile.GFile(PATH_TO_CKPT, 'rb') as fid: serialized_graph = fid.read() od_graph_def.ParseFromString(serialized_graph) tf.import_graph_def(od_graph_def, name='') sess = tf.Session(graph=detection_graph) print("> ====== Hand Inference graph loaded.") ``` - Detect hands. In this repo, this is done in the `utils/detector_utils.py` script by the `detect_objects` method. ```python (boxes, scores, classes, num) = sess.run( [detection_boxes, detection_scores, detection_classes, num_detections], feed_dict={image_tensor: image_np_expanded}) ``` - Visualize detected bounding detection_boxes. In this repo, this is done in the `utils/detector_utils.py` script by the `draw_box_on_image` method. This repo contains two scripts that tie all these steps together. - detect_multi_threaded.py : A threaded implementation for reading camera video input detection and detecting. Takes a set of command line flags to set parameters such as `--display` (visualize detections), image parameters `--width` and `--height`, videe `--source` (0 for camera) etc. - detect_single_threaded.py : Same as above, but single threaded. This script works for video files by setting the video source parameter videe `--source` (path to a video file). ```cmd # load and run detection on video at path "videos/chess.mov" python detect_single_threaded.py --source videos/chess.mov ``` > Update: If you do have errors loading the frozen inference graph in this repo, feel free to generate a new graph that fits your TF version from the model-checkpoint in this repo. Use the [export_inference_graph.py](https://github.com/tensorflow/models/blob/master/research/object_detection/export_inference_graph.py) script provided in the tensorflow object detection api repo. More guidance on this [here](https://pythonprogramming.net/testing-custom-object-detector-tensorflow-object-detection-api-tutorial/?completed=/training-custom-objects-tensorflow-object-detection-api-tutorial/). ## Thoughts on Optimization. A few things that led to noticeable performance increases. - Threading: Turns out that reading images from a webcam is a heavy I/O event and if run on the main application thread can slow down the program. I implemented some good ideas from [Adrian Rosebuck](https://www.pyimagesearch.com/2017/02/06/faster-video-file-fps-with-cv2-videocapture-and-opencv/) on parrallelizing image capture across multiple worker threads. This mostly led to an FPS increase of about 5 points. - For those new to Opencv, images from the `cv2.read()` method return images in [BGR format](https://www.learnopencv.com/why-does-opencv-use-bgr-color-format/). Ensure you convert to RGB before detection (accuracy will be much reduced if you dont). ```python cv2.cvtColor(image_np, cv2.COLOR_BGR2RGB) ``` - Keeping your input image small will increase fps without any significant accuracy drop.(I used about 320 x 240 compared to the 1280 x 720 which my webcam provides). - Model Quantization. Moving from the current 32 bit to 8 bit can achieve up to 4x reduction in memory required to load and store models. One way to further speed up this model is to explore the use of [8-bit fixed point quantization](https://heartbeat.fritz.ai/8-bit-quantization-and-tensorflow-lite-speeding-up-mobile-inference-with-low-precision-a882dfcafbbd). Performance can also be increased by a clever combination of tracking algorithms with the already decent detection and this is something I am still experimenting with. Have ideas for optimizing better, please share! <img src="images/general.jpg" width="100%"> Note: The detector does reflect some limitations associated with the training set. This includes non-egocentric viewpoints, very noisy backgrounds (e.g in a sea of hands) and sometimes skin tone. There is opportunity to improve these with additional data. ## Integrating Multiple DNNs. One way to make things more interesting is to integrate our new knowledge of where "hands" are with other detectors trained to recognize other objects. Unfortunately, while our hand detector can in fact detect hands, it cannot detect other objects (a factor or how it is trained). To create a detector that classifies multiple different objects would mean a long involved process of assembling datasets for each class and a lengthy training process. > Given the above, a potential strategy is to explore structures that allow us **efficiently** interleave output form multiple pretrained models for various object classes and have them detect multiple objects on a single image. An example of this is with my primary use case where I am interested in understanding the position of objects on a table with respect to hands on same table. I am currently doing some work on a threaded application that loads multiple detectors and outputs bounding boxes on a single image. More on this soon.

A list of practical projects that anyone can solve in any programming language (See solutions). These projects are divided into multiple categories, and each category has its own folder. To get started, simply fork this repo. CONTRIBUTING See ways of contributing to this repo. You can contribute solutions (will be published in this repo) to existing problems, add new projects, or remove existing ones. Make sure you follow all instructions properly. Solutions You can find implementations of these projects in many other languages by other users in this repo. Credits Problems are motivated by the ones shared at: Martyr2’s Mega Project List Rosetta Code Table of Contents Numbers Classic Algorithms Graph Data Structures Text Networking Classes Threading Web Files Databases Graphics and Multimedia Security Numbers Find PI to the Nth Digit - Enter a number and have the program generate PI up to that many decimal places. Keep a limit to how far the program will go. Find e to the Nth Digit - Just like the previous problem, but with e instead of PI. Enter a number and have the program generate e up to that many decimal places. Keep a limit to how far the program will go. Fibonacci Sequence - Enter a number and have the program generate the Fibonacci sequence to that number or to the Nth number. Prime Factorization - Have the user enter a number and find all Prime Factors (if there are any) and display them. Next Prime Number - Have the program find prime numbers until the user chooses to stop asking for the next one. Find Cost of Tile to Cover W x H Floor - Calculate the total cost of the tile it would take to cover a floor plan of width and height, using a cost entered by the user. Mortgage Calculator - Calculate the monthly payments of a fixed-term mortgage over given Nth terms at a given interest rate. Also, figure out how long it will take the user to pay back the loan. For added complexity, add an option for users to select the compounding interval (Monthly, Weekly, Daily, Continually). Change Return Program - The user enters a cost and then the amount of money given. The program will figure out the change and the number of quarters, dimes, nickels, pennies needed for the change. Binary to Decimal and Back Converter - Develop a converter to convert a decimal number to binary or a binary number to its decimal equivalent. Calculator - A simple calculator to do basic operators. Make it a scientific calculator for added complexity. Unit Converter (temp, currency, volume, mass, and more) - Converts various units between one another. The user enters the type of unit being entered, the type of unit they want to convert to, and then the value. The program will then make the conversion. Alarm Clock - A simple clock where it plays a sound after X number of minutes/seconds or at a particular time. Distance Between Two Cities - Calculates the distance between two cities and allows the user to specify a unit of distance. This program may require finding coordinates for the cities like latitude and longitude. Credit Card Validator - Takes in a credit card number from a common credit card vendor (Visa, MasterCard, American Express, Discoverer) and validates it to make sure that it is a valid number (look into how credit cards use a checksum). Tax Calculator - Asks the user to enter a cost and either a country or state tax. It then returns the tax plus the total cost with tax. Factorial Finder - The Factorial of a positive integer, n, is defined as the product of the sequence n, n-1, n-2, ...1, and the factorial of zero, 0, is defined as being 1. Solve this using both loops and recursion. Complex Number Algebra - Show addition, multiplication, negation, and inversion of complex numbers in separate functions. (Subtraction and division operations can be made with pairs of these operations.) Print the results for each operation tested. Happy Numbers - A happy number is defined by the following process. Starting with any positive integer, replace the number by the sum of the squares of its digits, and repeat the process until the number equals 1 (where it will stay), or it loops endlessly in a cycle which does not include 1. Those numbers for which this process ends in 1 are happy numbers, while those that do not end in 1 are unhappy numbers. Display an example of your output here. Find the first 8 happy numbers. Number Names - Show how to spell out a number in English. You can use a preexisting implementation or roll your own, but you should support inputs up to at least one million (or the maximum value of your language's default bounded integer type if that's less). Optional: Support for inputs other than positive integers (like zero, negative integers, and floating-point numbers). Coin Flip Simulation - Write some code that simulates flipping a single coin however many times the user decides. The code should record the outcomes and count the number of tails and heads. Limit Calculator - Ask the user to enter f(x) and the limit value, then return the value of the limit statement Optional: Make the calculator capable of supporting infinite limits. Fast Exponentiation - Ask the user to enter 2 integers a and b and output a^b (i.e. pow(a,b)) in O(LG n) time complexity. Classic Algorithms Collatz Conjecture - Start with a number n > 1. Find the number of steps it takes to reach one using the following process: If n is even, divide it by 2. If n is odd, multiply it by 3 and add 1. Sorting - Implement two types of sorting algorithms: Merge sort and bubble sort. Closest pair problem - The closest pair of points problem or closest pair problem is a problem of computational geometry: given n points in metric space, find a pair of points with the smallest distance between them. Sieve of Eratosthenes - The sieve of Eratosthenes is one of the most efficient ways to find all of the smaller primes (below 10 million or so). Graph Graph from links - Create a program that will create a graph or network from a series of links. Eulerian Path - Create a program that will take as an input a graph and output either an Eulerian path or an Eulerian cycle, or state that it is not possible. An Eulerian path starts at one node and traverses every edge of a graph through every node and finishes at another node. An Eulerian cycle is an eulerian Path that starts and finishes at the same node. Connected Graph - Create a program that takes a graph as an input and outputs whether every node is connected or not. Dijkstra’s Algorithm - Create a program that finds the shortest path through a graph using its edges. Minimum Spanning Tree - Create a program that takes a connected, undirected graph with weights and outputs the minimum spanning tree of the graph i.e., a subgraph that is a tree, contains all the vertices, and the sum of its weights is the least possible. Data Structures Inverted index - An Inverted Index is a data structure used to create full-text search. Given a set of text files, implement a program to create an inverted index. Also, create a user interface to do a search using that inverted index which returns a list of files that contain the query term/terms. The search index can be in memory. Text Fizz Buzz - Write a program that prints the numbers from 1 to 100. But for multiples of three print “Fizz” instead of the number and for the multiples of five print “Buzz”. For numbers which are multiples of both three and five print “FizzBuzz”. Reverse a String - Enter a string and the program will reverse it and print it out. Pig Latin - Pig Latin is a game of alterations played in the English language game. To create the Pig Latin form of an English word the initial consonant sound is transposed to the end of the word and an ay is affixed (Ex.: "banana" would yield anana-bay). Read Wikipedia for more information on rules. Count Vowels - Enter a string and the program counts the number of vowels in the text. For added complexity have it report a sum of each vowel found. Check if Palindrome - Checks if the string entered by the user is a palindrome. That is that it reads the same forwards as backward like “racecar” Count Words in a String - Counts the number of individual words in a string. For added complexity read these strings in from a text file and generate a summary. Text Editor - Notepad-style application that can open, edit, and save text documents. Optional: Add syntax highlighting and other features. RSS Feed Creator - Given a link to RSS/Atom Feed, get all posts and display them. Quote Tracker (market symbols etc) - A program that can go out and check the current value of stocks for a list of symbols entered by the user. The user can set how often the stocks are checked. For CLI, show whether the stock has moved up or down. Optional: If GUI, the program can show green up and red down arrows to show which direction the stock value has moved. Guestbook / Journal - A simple application that allows people to add comments or write journal entries. It can allow comments or not and timestamps for all entries. Could also be made into a shoutbox. Optional: Deploy it on Google App Engine or Heroku or any other PaaS (if possible, of course). Vigenere / Vernam / Ceasar Ciphers - Functions for encrypting and decrypting data messages. Then send them to a friend. Regex Query Tool - A tool that allows the user to enter a text string and then in a separate control enter a regex pattern. It will run the regular expression against the source text and return any matches or flag errors in the regular expression. Networking FTP Program - A file transfer program that can transfer files back and forth from a remote web sever. Bandwidth Monitor - A small utility program that tracks how much data you have uploaded and downloaded from the net during the course of your current online session. See if you can find out what periods of the day you use more and less and generate a report or graph that shows it. Port Scanner - Enter an IP address and a port range where the program will then attempt to find open ports on the given computer by connecting to each of them. On any successful connections mark the port as open. Mail Checker (POP3 / IMAP) - The user enters various account information include web server and IP, protocol type (POP3 or IMAP), and the application will check for email at a given interval. Country from IP Lookup - Enter an IP address and find the country that IP is registered in. Optional: Find the Ip automatically. Whois Search Tool - Enter an IP or host address and have it look it up through whois and return the results to you. Site Checker with Time Scheduling - An application that attempts to connect to a website or server every so many minute or a given time and check if it is up. If it is down, it will notify you by email or by posting a notice on the screen. Classes Product Inventory Project - Create an application that manages an inventory of products. Create a product class that has a price, id, and quantity on hand. Then create an inventory class that keeps track of various products and can sum up the inventory value. Airline / Hotel Reservation System - Create a reservation system that books airline seats or hotel rooms. It charges various rates for particular sections of the plane or hotel. For example, first class is going to cost more than a coach. Hotel rooms have penthouse suites which cost more. Keep track of when rooms will be available and can be scheduled. Company Manager - Create a hierarchy of classes - abstract class Employee and subclasses HourlyEmployee, SalariedEmployee, Manager, and Executive. Everyone's pay is calculated differently, research a bit about it. After you've established an employee hierarchy, create a Company class that allows you to manage the employees. You should be able to hire, fire, and raise employees. Bank Account Manager - Create a class called Account which will be an abstract class for three other classes called CheckingAccount, SavingsAccount, and BusinessAccount. Manage credits and debits from these accounts through an ATM-style program. Patient / Doctor Scheduler - Create a patient class and a doctor class. Have a doctor that can handle multiple patients and set up a scheduling program where a doctor can only handle 16 patients during an 8 hr workday. Recipe Creator and Manager - Create a recipe class with ingredients and put them in a recipe manager program that organizes them into categories like desserts, main courses, or by ingredients like chicken, beef, soups, pies, etc. Image Gallery - Create an image abstract class and then a class that inherits from it for each image type. Put them in a program that displays them in a gallery-style format for viewing. Shape Area and Perimeter Classes - Create an abstract class called Shape and then inherit from it other shapes like diamond, rectangle, circle, triangle, etc. Then have each class override the area and perimeter functionality to handle each shape type. Flower Shop Ordering To Go - Create a flower shop application that deals in flower objects and use those flower objects in a bouquet object which can then be sold. Keep track of the number of objects and when you may need to order more. Family Tree Creator - Create a class called Person which will have a name, when they were born, and when (and if) they died. Allow the user to create these Person classes and put them into a family tree structure. Print out the tree to the screen. Threading Create A Progress Bar for Downloads - Create a progress bar for applications that can keep track of a download in progress. The progress bar will be on a separate thread and will communicate with the main thread using delegates. Bulk Thumbnail Creator - Picture processing can take a bit of time for some transformations. Especially if the image is large. Create an image program that can take hundreds of images and converts them to a specified size in the background thread while you do other things. For added complexity, have one thread handling re-sizing, have another bulk renaming of thumbnails, etc. Web Page Scraper - Create an application that connects to a site and pulls out all links, or images, and saves them to a list. Optional: Organize the indexed content and don’t allow duplicates. Have it put the results into an easily searchable index file. Online White Board - Create an application that allows you to draw pictures, write notes and use various colors to flesh out ideas for projects. Optional: Add a feature to invite friends to collaborate on a whiteboard online. Get Atomic Time from Internet Clock - This program will get the true atomic time from an atomic time clock on the Internet. Use any one of the atomic clocks returned by a simple Google search. Fetch Current Weather - Get the current weather for a given zip/postal code. Optional: Try locating the user automatically. Scheduled Auto Login and Action - Make an application that logs into a given site on a schedule and invokes a certain action and then logs out. This can be useful for checking webmail, posting regular content, or getting info for other applications and saving it to your computer. E-Card Generator - Make a site that allows people to generate their own little e-cards and send them to other people. Do not use Flash. Use a picture library and perhaps insightful mottos or quotes. Content Management System - Create a content management system (CMS) like Joomla, Drupal, PHP Nuke, etc. Start small. Optional: Allow for the addition of modules/addons. Web Board (Forum) - Create a forum for you and your buddies to post, administer and share thoughts and ideas. CAPTCHA Maker - Ever see those images with letters numbers when you signup for a service and then ask you to enter what you see? It keeps web bots from automatically signing up and spamming. Try creating one yourself for online forms. Files Quiz Maker - Make an application that takes various questions from a file, picked randomly, and puts together a quiz for students. Each quiz can be different and then reads a key to grade the quizzes. Sort Excel/CSV File Utility - Reads a file of records, sorts them, and then writes them back to the file. Allow the user to choose various sort style and sorting based on a particular field. Create Zip File Maker - The user enters various files from different directories and the program zips them up into a zip file. Optional: Apply actual compression to the files. Start with Huffman Algorithm. PDF Generator - An application that can read in a text file, HTML file, or some other file and generates a PDF file out of it. Great for a web-based service where the user uploads the file and the program returns a PDF of the file. Optional: Deploy on GAE or Heroku if possible. Mp3 Tagger - Modify and add ID3v1 tags to MP3 files. See if you can also add in the album art into the MP3 file’s header as well as other ID3v2 tags. Code Snippet Manager - Another utility program that allows coders to put in functions, classes, or other tidbits to save for use later. Organized by the type of snippet or language the coder can quickly lookup code. Optional: For extra practice try adding syntax highlighting based on the language. Databases SQL Query Analyzer - A utility application in which a user can enter a query and have it run against a local database and look for ways to make it more efficient. Remote SQL Tool - A utility that can execute queries on remote servers from your local computer across the Internet. It should take in a remote host, user name, and password, run the query and return the results. Report Generator - Create a utility that generates a report based on some tables in a database. Generates sales reports based on the order/order details tables or sums up the day's current database activity. Event Scheduler and Calendar - Make an application that allows the user to enter a date and time of an event, event notes, and then schedule those events on a calendar. The user can then browse the calendar or search the calendar for specific events. Optional: Allow the application to create re-occurrence events that reoccur every day, week, month, year, etc. Budget Tracker - Write an application that keeps track of a household’s budget. The user can add expenses, income, and recurring costs to find out how much they are saving or losing over a period of time. Optional: Allow the user to specify a date range and see the net flow of money in and out of the house budget for that time period. TV Show Tracker - Got a favorite show you don’t want to miss? Don’t have a PVR or want to be able to find the show to then PVR it later? Make an application that can search various online TV Guide sites, locate the shows/times/channels and add them to a database application. The database/website then can send you email reminders that a show is about to start and which channel it will be on. Travel Planner System - Make a system that allows users to put together their own little travel itinerary and keep track of the airline/hotel arrangements, points of interest, budget, and schedule. Graphics and Multimedia Slide Show - Make an application that shows various pictures in a slide show format. Optional: Try adding various effects like fade in/out, star wipe, and window blinds transitions. Stream Video from Online - Try to create your own online streaming video player. Mp3 Player - A simple program for playing your favorite music files. Add features you think are missing from your favorite music player. Watermarking Application - Have some pictures you want copyright protected? Add your own logo or text lightly across the background so that no one can simply steal your graphics off your site. Make a program that will add this watermark to the picture. Optional: Use threading to process multiple images simultaneously. Turtle Graphics - This is a common project where you create a floor of 20 x 20 squares. Using various commands you tell a turtle to draw a line on the floor. You have moved forward, left or right, lift or drop the pen, etc. Do a search online for "Turtle Graphics" for more information. Optional: Allow the program to read in the list of commands from a file. GIF Creator A program that puts together multiple images (PNGs, JPGs, TIFFs) to make a smooth GIF that can be exported. Optional: Make the program convert small video files to GIFs as well. Security Caesar cipher - Implement a Caesar cipher, both encoding, and decoding. The key is an integer from 1 to 25. This cipher rotates the letters of the alphabet (A to Z). The encoding replaces each letter with the 1st to 25th next letter in the alphabet (wrapping Z to A). So key 2 encrypts "HI" to "JK", but key 20 encrypts "HI" to "BC". This simple "monoalphabetic substitution cipher" provides almost no security, because an attacker who has the encoded message can either use frequency analysis to guess the key, or just try all 25 keys.

# AD-Prediction Convolutional Neural Networks for Alzheimer's Disease Prediction Using Brain MRI Image ## Abstract Alzheimers disease (AD) is characterized by severe memory loss and cognitive impairment. It associates with significant brain structure changes, which can be measured by magnetic resonance imaging (MRI) scan. The observable preclinical structure changes provides an opportunity for AD early detection using image classification tools, like convolutional neural network (CNN). However, currently most AD related studies were limited by sample size. Finding an efficient way to train image classifier on limited data is critical. In our project, we explored different transfer-learning methods based on CNN for AD prediction brain structure MRI image. We find that both pretrained 2D AlexNet with 2D-representation method and simple neural network with pretrained 3D autoencoder improved the prediction performance comparing to a deep CNN trained from scratch. The pretrained 2D AlexNet performed even better (**86%**) than the 3D CNN with autoencoder (**77%**). ## Method #### 1. Data In this project, we used public brain MRI data from **Alzheimers Disease Neuroimaging Initiative (ADNI)** Study. ADNI is an ongoing, multicenter cohort study, started from 2004. It focuses on understanding the diagnostic and predictive value of Alzheimers disease specific biomarkers. The ADNI study has three phases: ADNI1, ADNI-GO, and ADNI2. Both ADNI1 and ADNI2 recruited new AD patients and normal control as research participants. Our data included a total of 686 structure MRI scans from both ADNI1 and ADNI2 phases, with 310 AD cases and 376 normal controls. We randomly derived the total sample into training dataset (n = 519), validation dataset (n = 100), and testing dataset (n = 67). #### 2. Image preprocessing Image preprocessing were conducted using Statistical Parametric Mapping (SPM) software, version 12. The original MRI scans were first skull-stripped and segmented using segmentation algorithm based on 6-tissue probability mapping and then normalized to the International Consortium for Brain Mapping template of European brains using affine registration. Other configuration includes: bias, noise, and global intensity normalization. The standard preprocessing process output 3D image files with an uniform size of 121x145x121. Skull-stripping and normalization ensured the comparability between images by transforming the original brain image into a standard image space, so that same brain substructures can be aligned at same image coordinates for different participants. Diluted or enhanced intensity was used to compensate the structure changes. the In our project, we used both whole brain (including both grey matter and white matter) and grey matter only. #### 3. AlexNet and Transfer Learning Convolutional Neural Networks (CNN) are very similar to ordinary Neural Networks. A CNN consists of an input and an output layer, as well as multiple hidden layers. The hidden layers are either convolutional, pooling or fully connected. ConvNet architectures make the explicit assumption that the inputs are images, which allows us to encode certain properties into the architecture. These then make the forward function more efficient to implement and vastly reduce the amount of parameters in the network. #### 3.1. AlexNet The net contains eight layers with weights; the first five are convolutional and the remaining three are fully connected. The overall architecture is shown in Figure 1. The output of the last fully-connected layer is fed to a 1000-way softmax which produces a distribution over the 1000 class labels. AlexNet maximizes the multinomial logistic regression objective, which is equivalent to maximizing the average across training cases of the log-probability of the correct label under the prediction distribution. The kernels of the second, fourth, and fifth convolutional layers are connected only to those kernel maps in the previous layer which reside on the same GPU (as shown in Figure1). The kernels of the third convolutional layer are connected to all kernel maps in the second layer. The neurons in the fully connected layers are connected to all neurons in the previous layer. Response-normalization layers follow the first and second convolutional layers. Max-pooling layers follow both response-normalization layers as well as the fifth convolutional layer. The ReLU non-linearity is applied to the output of every convolutional and fully-connected layer.  The first convolutional layer filters the 224x224x3 input image with 96 kernels of size 11x11x3 with a stride of 4 pixels (this is the distance between the receptive field centers of neighboring neurons in a kernel map). The second convolutional layer takes as input the (response-normalized and pooled) output of the first convolutional layer and filters it with 256 kernels of size 5x5x48. The third, fourth, and fifth convolutional layers are connected to one another without any intervening pooling or normalization layers. The third convolutional layer has 384 kernels of size 3x3x256 connected to the (normalized, pooled) outputs of the second convolutional layer. The fourth convolutional layer has 384 kernels of size 3x3x192 , and the fifth convolutional layer has 256 kernels of size 3x3x192. The fully-connected layers have 4096 neurons each. #### 3.2. Transfer Learning Training an entire Convolutional Network from scratch (with random initialization) is impractical[14] because it is relatively rare to have a dataset of sufficient size. An alternative is to pretrain a Conv-Net on a very large dataset (e.g. ImageNet), and then use the ConvNet either as an initialization or a fixed feature extractor for the task of interest. Typically, there are three major transfer learning scenarios: **ConvNet as fixed feature extractor:** We can take a ConvNet pretrained on ImageNet, and remove the last fully-connected layer, then treat the rest structure as a fixed feature extractor for the target dataset. In AlexNet, this would be a 4096-D vector. Usually, we call these features as CNN codes. Once we get these features, we can train a linear classifier (e.g. linear SVM or Softmax classifier) for our target dataset. **Fine-tuning the ConvNet:** Another idea is not only replace the last fully-connected layer in the classifier, but to also fine-tune the parameters of the pretrained network. Due to overfitting concerns, we can only fine-tune some higher-level part of the network. This suggestion is motivated by the observation that earlier features in a ConvNet contains more generic features (e.g. edge detectors or color blob detectors) that can be useful for many kind of tasks. But the later layer of the network becomes progressively more specific to the details of the classes contained in the original dataset. **Pretrained models:** The released pretrained model is usually the final ConvNet checkpoint. So it is common to see people use the network for fine-tuning. #### 4. 3D Autoencoder and Convolutional Neural Network We take a two-stage approach where we first train a 3D sparse autoencoder to learn filters for convolution operations, and then build a convolutional neural network whose first layer uses the filters learned with the autoencoder.  #### 4.1. Sparse Autoencoder An autoencoder is a 3-layer neural network that is used to extract features from an input such as an image. Sparse representations can provide a simple interpretation of the input data in terms of a small number of \parts by extracting the structure hidden in the data. The autoencoder has an input layer, a hidden layer and an output layer, and the input and output layers have same number of units, while the hidden layer contains more units for a sparse and overcomplete representation. The encoder function maps input x to representation h, and the decoder function maps the representation h to the output x. In our problem, we extract 3D patches from scans as the input to the network. The decoder function aims to reconstruct the input form the hidden representation h. #### 4.2. 3D Convolutional Neural Network Training the 3D convolutional neural network(CNN) is the second stage. The CNN we use in this project has one convolutional layer, one pooling layer, two linear layers, and finally a log softmax layer. After training the sparse autoencoder, we take the weights and biases of the encoder from trained model, and use them a 3D filter of a 3D convolutional layer of the 1-layer convolutional neural network. Figure 2 shows the architecture of the network. #### 5. Tools In this project, we used Nibabel for MRI image processing and PyTorch Neural Networks implementation.

🚀 This is a framework for Android application development that incorporates the author's years of industry experience in Android. It is designed to efficiently develop and build Android mobile apps. Even beginners who are new to Android application development can quickly get started with it.