I host live office hours on this page every Thursday from 2-3PM EDT (you can convert that time to your own timezone here).
Office hours are totally free and anyone is welcome to join in and view the live office hours feed.
During office hours I answer questions I have received from PyImageSearch readers just like yourself over the week. For each question I answer I donate $50 of my own personal funds to a selected charity of the week.
There will be no Office Hours this coming Thursday (Sept. 19th).
Myself and the PyImageSearch team just released the Hobbyist Bundle of Raspberry Pi for Computer Vision this morning (Wednesday, Sept. 18th), and to be totally candid, I’m feeling a bit burnt out and exhausted.
I need to use the time spent during Office Hours to recharge my batteries a bit.
Questions will be added ~24 hours before each Office Hours session.
Tune in on this page to view the live office hours broadcast!
Office Hours Archives
Below you’ll find archives of previous office hour sessions. These sessions are free to view.
Office Hours - 2019 (2)
In our very first office hours session we discuss:
- (11:40) Why are you offering office hours?
- What is the goal of providing office hours?
- (12:15) My advice to PhD and MSc students when finding a research topic
- How to find a good advisor
- How to find an advisor whose research interests align with your own
- How to successfully get through graduate school
- (21:30) Array slicing and cropping ROIs in images
- How to manually find (x, y)-coordinates for a given object
- Accessing individual pixel values
- NumPy array slicing syntax
- (29:30) Array slicing and cropping ROIs in video
- Quick hack to export single frames for manual inspection
- (35:45) Image hashing and face recognition/verification
- Is image hashing the correct algorithm to use?
- What might better alternatives be?
- (39:55) Building a product search engine
- How might we build an image search engine for images of shoes?
- What types of algorithms should we use?
- And how can these algorithms be made searchable?
- Is data augmentation appropriate for CBIR?
- Is data augmentation a good idea when building an image search engine?
Charity of the week:
- Prostate Cancer Foundation (PCF)
- Amount donated: $300
Office Hours slides:
- (09:15) Command line arguments and argparse
- Learn how to properly use command line arguments
- Resolve common “USAGE” errors that you will encounter
- (20:50) What IDEs do you recommend for computer vision and deep learning development?
- Sublime Text for smaller projects
- PyCharm for larger projects
- For RPi try to use remote development
- (23:20) How can we use OpenCV OCR for newspaper text recognition?
- (27:20) Hand gesture recognition on the RPi
- Can hand gesture recognition run in real-time on the RPi?
- What types of models should we use?
- (31:20) How can we improve face recognition accuracy?
- Consider your face detector
- Utilize a combination of facial embeddings and machine learning
- (38:30) How can we build a web crawler for images?
- Check out Scrapy
- Useful for building an image search engine
- (41:12) How can I parse object detection annotations?
- (44:50) Producer/consumer with Python
- How can we use a producer/consumer relationship to build a faster, more efficient deep learning inference system?
- (48:10) How can we perform object detection with OpenCV?
- What do the returned values actually mean?
- How can we take them and convert them to bounding box coordinates and labels?
- (53:20) Threads vs. processes
- When do we use threads?
- And how can processes be used to improve program speed?
- (55:20) How can we do object detection with multiple cameras?
- Use multiple cameras but only a single model
- Use producer/consumer relationship to keep queue filled
- Point of diminishing returns — there is a limit to the number of cameras you can attach to a single computer
Charity of the week:
- Beagle Rescue of Southern Maryland
- Amount donated: $450
Office Hours slides:
Office Hours FAQ
Below are are answers to common questions I get asked regarding office hours.
Why office hours?
5 years ago I started PyImageSearch with the goal of empowering students, developers, and researchers with free tutorials and free help to master computer vision, deep learning, and OpenCV. I believe, on all counts, that the journey has been a huge success.
I love running PyImageSearch, because after many years, I’ve found that what gives me energy is helping others. Helping someone learn a new concept or algorithm, and have it just *click* and the lightbulb flip *on* is what motivates me to teach and help.
But there’s a problem…it’s hard to help everyone.
On an average day I receive ~200 emails and questions to my PyImageSearch inbox.
While I enjoy taking the time to answer questions, trying to handle that load on a daily basis just isn’t feasible.
I ran the numbers (because I’m a nerd and I like numbers) and calculated that I could spend ~71.8% of my week exclusively answering reader questions, but, to be blunt, the electric company doesn’t let me pay my bill with “how many questions I answered today” (although that would be really awesome if they did).
Email-based questions are especially hard because it’s 1:1 help…no one else benefits from the discussion. Facebook Messenger, Twitter DMs, and LinkedIn messages are the same. 1:1 help is just not an efficient use of our time, especially since my goal is to “do the most good” and help as many people in the computer vision/deep learning community as a whole.
By doing office hours I can now:
- Efficiently answer your questions
- Share the knowledge with the rest of the community so everyone can benefit
Furthermore, my Dad being diagnosed with cancer last year has been an eye opening experience for me. Selfishly, I want to spend as much time with him as I possibly can — and to that end, I am taking action in both my business and personal life to facilitate the time left with my family.
I believe office hours are the best possible solution as:
- The community still receives free help
- Charity benefits (for each question I answer I donate $50 of my own personal funds to the specified charity of the week)
- And I reclaim time with my loved ones
It’s win/win all the way around.
When are office hours?
Office hours are held every week on Thursday from 2-3PM EDT (unless otherwise noted). You can convert that time to your own timezone here.
Where can I watch the live office hours video feed?
When office hours start this page will be updated with a live video stream. I use Facebook Live for the office hours feed, so along with monitoring this page, you can also follow PyImageSearch on Facebook to be notified when the office hours stream goes live.
How can I get my question answered?
If you have a question that you would like me to answer you can either (1) email it to me or (2) use my contact form to send me a message.
Kindly understand that I receive over 200+ emails per day, and while I wish I could answer each and every incoming question, there’s only so much time in the day — I cannot get to them all (but I do try). Please don’t get offended if you ask a question and I cannot get to it.
I bought one of your books/courses. How can I get my questions answered?
As I mention in my FAQ, if you are my customer you receive priority support. I will of course continue to provide you with help and support (that will never change). Just send me an email with your question and include your purchase reference. I’ll then be able to get back to you.
Are your office hours free?
Yes, my office hours are 100% free. Anyone, anywhere in the world can view and learn from my office hours.
Are your office hours recorded?
I understand that due to timezone differences and life obligations that not everyone can make it to the live office hours.
But don’t worry!
I record every office hours session and will update this page with the office hour archives, that way you can watch the office hours later, even if you cannot make it to the live session.
Why do you donate to charity?
For each question that I answer during a live office hours session I donate $50 to the specified charity of the week.
I’m a big believer in the phrase “do the most good”. When I first started PyImageSearch I was a graduate student wrapping up my PhD dissertation. I donated my time to helping others for free.
PyImageSearch has grown tremendously over the past 5 years, and coupled with my own personal obligations, I have less time than I used to.
Since I cannot donate my time like I used to, I now donate my own personal funds, simply because I believe in doing the most good for the community.