In this assignment, you will implement recurrent neural networks and apply them to image captioning on the Microsoft COCO data. You will also explore methods for visualizing the features of a pretrained model on ImageNet, and use this model to implement Style Transfer. Finally, you will train a Generative Adversarial Network to generate images that look like a training dataset!
Assignment 3.pdf
Download data. Next, you will need to download the COCO captioning data, a pretrained SqueezeNet model (for TensorFlow), and a few ImageNet validation images. Run the following from the assignment3 directory:
1. If you selected Option A and worked on the assignment in Colab, open collect_submission.ipynb in Colab and execute the notebook cells. If you selected Option B and worked on the assignment locally, run the bash script in assignment3 by executing bash collectSubmission.sh.
FRI DEC 12finals study guide txtbombs source code directoryMON DEC 8gdb notes pdftxtFRI DEC 5bomblab (ass'n 11) pdfbombs directorydue fri dec 12WED DEC 3adv cpu topics txtpls rtn kit on friday! pdfpls do course eval! htmlMON DEC 1assignment 10: caches pdf due fri dec 5FRI NOV 14assignment 9: pipe optimization pdf files tarFRI NOV 7assignment 8: cpu research pdfTUE NOV 4take-home midterm due Wed Nov 5 pdfMON NOV 3assignment 7 practice pdfFRI OCT 31assignment 7: y86 and seq pdfdue Nov 7 assignment7.tar tar ass'n 5 back ass'n 6 back ass'n 6 answers ISA slides pdf SEQ slides pdfWED OCT 22midterm (open data sheets, ass'n 6 not included)MON OCT 20midterm review notes txtWED OCT 15assignment 6b: mapping a RAM pdf due Oct 20 (along with Ass'n 6) 6.270 hardware pdf Cypress 6264 8K RAM pdf Handy Board schem gif 68HC11 big manual pdfFRI OCT 10assignment 6: address decoding pdf due Oct 17 expanddemo.s mystery5000.relMON OCT 6assignment 5 pdf due Oct 15 README FIRST! BootLoad.java ReadRel.java Serial.java serialxmit.s BootTerm.java BootTermSafe.java analogdemo.sMON SEP 29intro to 6811 pdfm68hc11e manual pdfassignment 4 pdfhtml due Oct 6 java setup html HC11Boot.java Serial.java beep.sWED SEP 24garage state machine pdfMON SEP 22logic implementation of eqns pdf tiffFRI SEP 19assignment 3 pdfhtml due Sep 29logic eqn reading from Fletcher no linkMON SEP 15schematic hints htmlpdfmystery hints pdfFRI SEP 12transistor reading from horowitz/hill no linkpp 117-141 from Tanenbaum no linkMON SEP 8assignment 2 pdfdue Sep 15data sheets htmluml305devhtmlpdfparts listpdfWED SEP 3assignment 1 pdfdue Sep 8 resources COURSE EVAL ikonboard software data links ASSIGNMENT 3: State Machinesout: Fri Sep 19due at the start of class, Mon Sep 29.
In this assignment, you will design and implement a state machine that acts as the controller for the line-following robot presented in class.You will have to demonstrate a working solution to the professor or TA. Please check the class web site for office hours and plan ahead!The steps are:1. Draw the State Transition Diagram showing each state, the outputs produced in each state, and the transition rules between states. This is already done (see below).2. From the State Transition Diagram, create a State Transition Table. The Table maps from inputs (current state bits and external input bits) to outputs (next state bits and control output bits). This too is already done (see below).3. From the State Transition Table, create Logic Equations that describe the Table’s output bits as a function of its input bits.4. Minimize these Logic Equations.5. Implement the simplified equations in discrete logic, using the chips provided in your kit.6. Build and test the resulting solution, using a latch chip to store the state bits and an external clock to run the machine.7. When the state machine appears to be working, connect it to the robot and see if it can follow the line!We will start this problem set by building an oscillator and dividing this frequency down to a point where it can be used in our state machine.PROBLEM 3-1. INVERTER CHAIN AS OSCILLATOR.The following circuit will produce an oscillator—a circuit that continually flips back and forth between 0 and 1:
Do not worry if the STEP problems appear difficult: they are meant to be! However, you should not be daunted. These assignments are designed to help you to develop the skills you need, over time, so that by the time you sit the STEP exam in the summer of Y13 you will feel well-prepared.
The assignment is published as a pdf file below. Each STEP Support assignment module starts with a warm-up exercise, followed by preparatory work leading to a STEP question. Finally, there is a warm-down exercise.
The warm up for this assignment involves the sigma notation, and a proof of the formula for the sum of the terms of a geometric progression. For the last part, use the formula rather than summing the individual terms, and try to do this without using a calculator.
More information on Diophantine equations can be found in this article on Plus, the free online mathematics magazine, and in this Wikipedia entry.You may also enjoy watching this talk by Dr Vicky Neale - 'How to Solve Equations'This Plus article shows how we can integrate from first principles.This Numberphile video discusses the "Monkeys and Coconuts" problem. Do not watch it until you have tried the assignment!
In the settings set Require students click submit button to No. Students can continue to make changes to their assignment and at no point do they 'submit'. If the work will be graded at some point it is recommended that either Prevent late submissions is set to Yes to ensure that no changes can be made after the due date, or all submissions are locked when grading commences to ensure that the work is not altered during grading.
In the settings, set "Students submit in groups" to Yes. If you just do this, then once one student has submitted, the assignment will be flagged as submitted even if the others haven't contributed. If you want to ensure everyone has an input, set "Require students click submit button" to Yes and then change "Require all group members to submit" to Yes. The assignment will only be classed as submitted when each member has contributed, and once one student has submitted, the remaining members's names will be displayed for the group to see who still needs to add their input.
In the settings, choose 'Anonymous submissions'. When students submit assignments, their names will be replaced by randomly-generated participant numbers so you will not know who is who. Note that this is not totally anonymous because you can reveal their identities in the assignment settings and you can work out identities from the logs - so this might not be suitable if your establishment has very precise privacy requirements.
In the settings, choose "Offline grading worksheet". When students have submitted, click "View/grade all submissions" and you can download their assignments from the link "Download all submissions" and download the grading sheet from the link "Download grading worksheet". You can then edit grades and re-upload the grading worksheet. You can also upload multiple feedback files in a zip from this drop down menu. See Assignment settings for an explanation of how to use the "upload multiple feedback files as zip" feature.
The first page students will see when they click on the assignment activity link from the course page will display the assignment name, description and the submission status. The first time a student views the assignment it will look like this:
Clicking 'Grade' will take you to the first student in the list so you can start grading individually. If you wish to grade several assignments, clicking Save and Show next will take you to the next submission.
If you need to divide submissions between more than one person, you can apply groups to the assignment and let markers know which group(s) to mark. Note that because group membership is not itself anonymised, this may make anonymised submissions that bit less anonymous, though as long as the groups aren't very small this should be acceptable.
To access the Quick Grading interface, from the Grading Summary page click 'View all submissions'; the Grading Table displays. Scroll to bottom of the page to configure Options, and check the box for 'Quick grading'. While you're down there, you can also set the number of assignments to display per page, filter the assignments e.g. to see who has not submitted, unmarked assignments, etc.
If you have enabled File Feedback in the Assignment settings and wish to upload either the marked student assignment, a completed text based feedback document or audio feedback, click on the green tick in the Grade column (or use the icon in the Edit column and select Grade).
Note: Helpfully that downloaded worksheet will contain any existing grades and summary comments which have already been given for that assignment i.e. if marking has already started. However, to see pre-existing comments fully you may need to set your spreadsheet to 'wrap text' within cells.
If you have high level feedback you want to give to an entire cohort, it is generally a good idea to give this feedback in the context of the assignment, rather than e.g. separately via a Forum. Moodle allows you to select some or all students and attach a single, common feedback file to their assignment feedback. This common feedback will appear to each student along with any other individual feedback files you have prepared for each. 2ff7e9595c
Comments