Ging’s Status Report for 09/30

In this week, I personally have prepared the slides for design review and practiced presenting it. I also found a dataset for our training. This is the link of the dataset I’ve found and tried: http://gibsonenv.stanford.edu/method/. The slides are in the design review tab.

My progress is on schedule. Good work!

Next week, I hope to give the presentation, and do the peer reviews. I hope to start collecting data and start training.

The ECE courses that covered the design is: 18794 computer vision, 18213 computer systems, 17313 software engineering. They are helpful for building a deep learning network and managing the project well.

Team Status Report for 09/30

The most significant risk that could jeopardize the success of the project is time management. We have many deadlines and we have to find time really working on the projects. The risks are managed by our schedule. If behind schedule, we will catch up on it.

One change is made: we added a part called Gibson’s Goggle in the block diagram. The change is necessary to pre-process the input signal/image data strea. No additional cost is involved.

No change in schedule.

An image of our updated diagram:

Team Status Report for 09/23

Our project aims to guide the blind people. When developing the project proposal, we have consideration of positive social factors as well as public health. The guidance recognition system, as a substitute for guide dogs, is a cheaper alternative for blind people. By combining the recognition system with walkers, we hope to bring dignity and convenience to blind people with more accessibility.  It is important to consider the effect of our project on society because as engineers, we should always build for social good and welfare. As an underrepresented group, blind people have the right to enjoy the advantages of high technology and modern life.

The risk that could jeopardize the project is whether we can follow the schedule. We are tracking our progress actively for scheduling purpose. This week we already ordered components from ECE. Continency plan is that we can postpone the work after MVP. No change to the design yet.

Weekly Specific Question Answer: As a project whose main users are the blind persons, we consider the safety of our project:  More specifically, how our product will perform when there are sudden change in environment that would make our product’s instructions not correct. Thus, we plan to have an emergency protocol. This is important because our clients are vulnerable to environment change and safety is at top priority of our clients.

Ging’s Status Report for 09/23

This week we spent lecture time watching other teams’ presentations of their project proposal and doing peer reviews. I learned a lot from other’s presentations and feedback. I realize that we should specify which feature the detection system is to measure and specify the details of the accuracy into a 2×2 grid. This is the training example I’ve found on github: https://github.com/devendrachaplot/Neural-SLAM.

This week I also did some research on object classification and image classification model to familiarize myself with computer vision models.

I’m on schedule.

Next week I will start implementing the SLAM model.