Problem Set 8

Exercises are taken from the textbook chapters. Maximum score for this problem set is 29 points [21 points for 6.4210] and the due date is Friday, November 8th, 2024 (11:59PM Eastern Time)

  • Please ensure your manipulation version is at least manipulation==2024.11.04.post1 when completing this pset. If working on Deepnote, please ensure the Init notebook has been run.
  • Exercise 9.1: (a) 3 points, (b) 1 point, (c) 1 point, (d) 1 point, (e) 2 points.
  • Exercise 9.2: (a) 4 points, (b) 2 points, (c) 2 points (d) 2 points.
  • Exercise 10.1 [6.4212 Only]: (a) 4 points, (b) 4 points.
  • Project Check-in 1: 2 points. Please briefly describe the infrastructure/simulation set-up that your team has created for the project. Some example details to include are: what simulator (i.e. Drake, PyBullet, MoJoCo, etc), what development pipeline (i.e. Deepnote, local installation, colab, etc), what robot (i.e. Kuka, Panda, UR5, etc), what objects or items have been added to the simulation environment (i.e. a block, bins, shelves, etc) and any other relevant details. Include a screenshot of your simulation environment. You should also meet with the assigned TA for your project to discuss your progress and any issues that you are facing. Please contact your TA to set up a time to meet, either during regular office hours or during special office hours (Nov 6, 9:30-11am in 45-230) or make special arrangements.
  • Survey: 1 point. After submitting the survey, write the keyword at the end of your written submission.

You will submit four files for this problem set (no need to submit 10.1 if you are in 6.4210):

  • Problem Set 8 written problems. All written problems (Exercise 9.1 (b, c, d, e), 9.2 (b, c, d)) belong in this submission, as well as the survey code and project update. Please submit one PDF for all written problems. Note that your solutions to written problems only need to be within the PDF and do not need to be duplicated in the notebooks. Please ensure that if you are submitted scans of handwritten solutions that the text is legible.
  • Problem Set 8 notebook problems. The final versions of your notebooks: label_generation.ipynb, segmentation_and_grasp.ipynb and contrastive.ipynb (downloaded from Deepnote). Please double check that the notebooks you are submitting have exactly these names.
When submitting to Gradescope, include all files in the same upload. This way, the autograder will be able to grade your notebooks, and the teaching staff will be able to read your written work. Please do not submit any other files. For the notebook problems, Gradescope will give you a grade. You can resubmit as many times as you like before the deadline.

Finally, we have a small survey. After you take the survey, you will be given a code. Include this code in your written submission.