I work as a Research and Development Engineer for a biomedical product company. My work mainly focuses on Embedded systems and Signal processing. I would be applying machine learning techniques to all sorts of electrophysiological signals for their classification.
My name is Hung. I’m software engineer and Research Fellow at Topica AI Lab. My main research are Computer Vision (ALPR) and Deep Learning. I teach kids to code and make Robotics project at Maker Hanoi, every weekend.
Nice to see you guys, here!
My linkedin Profile:
Hi. Hafidz here from Malaysia. I have previously in the telco domain as a data scientist and have now moved to a job marketplace space. I’m excited to be part of this course to learn more about practical applications of deep learning, as I’ve only worked with ML (customer segmentation, classification, forecasting) problems in the telecom domain (in my previous company). My dream is to start an AI company dealing with image generation and/or NLP at it’s core.
Hello from Omaha, Nebraska!
I grew up in Japan and moved to US after graduating from high school to check out this part of the world. I started out as Electrical Engineering major but switched to Computer Science after chatting with a friend who was writing a tic tac toe program for his CS class. I knew how I would play - I would pick the middle cell first if it was available, etc. But I could not quite explain why. So I thought “how would I teach a computer how to play when I just ‘know’?” That’s how it all started.
I’ve always been interested in Machine Learning, but it was not till June this year that I started to actually learn. I took Neural Networks for Machine Learning in Coursera, but after completing it, I did not feel ready to start building anything. Then I heard about fast.ai. I cannot tell you how happy I was to make my very first submission to a Kaggle competition right after completing the lesson 1 of the part 1 (v1)!!
My first reason of interest in deep learning is for its potential for Agritech applications: in summer
2016, I won a scholarship at WebValley, a data science summer school for high school students
run by Fondazione Bruno Kessler (FBK), that year dedicated to developing predictive classifiers
of fruit maturation by means of personal spectroscopy, imaging and deep learning models. My
background at 18y was that of a “magnet” type high school, with good foundations in Python
and databases, so it was great learning using Keras with TensorFlow and apply CNNs first
trained on ImageNet and then in transfer learning mode to some thousands of images from the
field or the production lines. I really enjoyed this experience of machine learning in Agritech
because I discovered that I can effectively combine my rural origins with the passion for
technology I have. Thus I applied for an internship at FBK while studying for a BSc in Computer
Science at Trento University, and I am now developing a deep learning model for grape quality
on paired FOSS (lab spectrometry) and portable spectrometry spectra, with sugar content and
other parameters as targets. I am learning to develop models in PyTorch and run them on local
GPUs and Azure cloud. I want to acquire skills on GANs to generate additional measures (these
are seasonal data) and develop models that can be run on the field for future real-time
applications of interest for wine production. I hope that the method will be really used by wine
I am Vishal Pandey From Patna, India… Currently I am doing my M.Tech In Analytics From NIT Rourkela . I have recently been introduced to deep learning . Hope to get most out of this course. Currently I am looking for some topic for my M.Tech thesis. I will be doing it in deep learning but right now have no idea what to work on. If anybody could suggest , it would be very kind helpful .
Good Luck to everyone on this wonderful journey on DEEP LEARNING with @jeremy and @rachel . Thanks to both of them for this exciting experience we r going to have… LOTS OF THANKS…
Since the last course I’ve been focused on the convergence between deep learning and recommender systems and I’m currently developing a system for real estate recommendation. Two of my coworkers have joined the course, one in person and one distance like me, and deep learning is rapidly becoming an important tool at Realtor.
I’m looking forward to learning pytorch and the new framework @jeremy is developing and I hope to contribute to the recommendation discussion when we reach that part of the course.
I’m from Lagos, Nigeria but currently based in Sweden working for one of the largest European Utilities. Looking to transition into an ML career and so I’ve been enrolled in the Udacity Nanodegree which I think is valuable in it’s own right. Currently working through the DL module of the course so it’s a great chance to be an international fellow in the DL course.
Hoping to take my learnings from this course and apply them initially to music analysis inspired by the Earworm series from Vox
Best part of the course so far is the huge number of Smaaaaaart folks in this fellowship from whom I hope to learn much as well as contribute too.
Let’s do this guys!!!
I am a software engineer. I discovered deep learning while building a Handwriting recognition system back in 2014 when I realised how ugly some of the conventional methods were. It was also about the same time when DL exploded
Took Andrew Ng’s course and enjoyed it a lot. Experimented with Theano but could not ‘port’ it to mobile phones. Took CS231n and CS224d. Early this year I joined Udacity’s Self Driving Car Nanodegree and it’s been a great learning experience.
I am experimenting with DL for Audio at work and also interested in image segmentation and multimodal learning for robotics. Amazed to be a part of this course.
Lets build some awesome projects together!
That’s amazing, thank you! I’m rarely at markets, but let Kirsten know that I said to give you the Deep Learning Discount next time.
There’s a fast.ai Meetup group starting in Victoria this Wednesday which might also interest you. I can send you the link if you’re interested!
I have a PhD in physics and was working as a researcher in experimental high energy physics and was teaching at the university. In my research I was also using machine learning algorithms like neural nets and boosted decision trees to search and measure rare decays. Recently I’ve decided to leave academia to take new challenges. Now, I’m working at a private company where I’m responsible for developing algorithms (using deep neural nets as well) to extract useful information from satellite images. So far I’ve checked several resources on the topic (i.e. Deep Learning Book, Stanford’s CS231n, articles, …) but I’m missing hands-on experience, which brought me here.
A software services company and a liquor store is an interesting combination
I saw an interesting article few days back in the same space that you are looking at http://www.businessinsider.com/walmart-store-robot-program-expands-2017-10
I think one can build a small robot (May be a webcam attached to a roomba or a Toy car ) and use some of the techniques that you will learn in the course to solve your problem.
I am a computer vision PhD student in Meknes, a small town in Morocco.
I work on medical imaging analysis. I’m looking for/developing efficient methods to segment
MRI images using few labels.
So I’m here to learn and understand quite well deep learning in order to come up with something new to solve my problem.
And also I want to have a broad view to spot other opportunities to apply this knowledge.
@alessiamarcolini This is a very interesting application. FBK seems to be doing a lot of good work…
I work at a Research lab in the areas of Machine Learning and Deep Learning. Over the years, I have worked on many aspects of Machine Learning like Machine Translation, Information Extraction, Stock Market prediction, Text Classification etc.
However, the most satisfying work so far has been a simple system that we developed for UNICEF to route text (SMS) messages to the right people. This was deployed by UNICEF to more than a dozen developing countries like Uganda, Sierra Leone, Syria, Zambia, etc. and apparently is a life saver in those communities.
For folks who are interested, here is an article that describes how it is being used in a refugee settlement in Uganda http://www.ureport.ug/story/188/
and a paper that describes the system https://pdfs.semanticscholar.org/768f/c32711e2562eeddc28a8015197b97e6fd90b.pdf
Just goes on to show how Machine Learning can make a huge social impact.
I took both part 1 and part 2 of the course and love @jeremy 's teaching style. I think I will be here for all the iterations of his courses
Hi, my name is Victor Arias, my background is in robotics, 3d printers, drones, etc. I’m from Colombia, and right now I live in Cucuta, Norte de Santander, Colombia. My primary goal is to take advantage of artificial intelligence techniques in order to offer health care products and services for a restrictive environment with low-cost medical mobile applications. For example, in this time we are working on the amount of calories calculation in a food´s dish; first, using a deep neural network to identify the food dish, and then creating 3D models of an image with the aim to assign calories and food composition. With this approach, you can remotely follow a patient´s diet with precision, i.e. in children with malnutrition or in obese people. An important stumbling block to overcome is the lack of internet access in rural areas of Colombia, so you cannot have the service tied to a cloud computing. In this regard, an alternative method is to design a few-parameters network for intelligent mobile applications. my hobby is to teach children how to program through video games and animations, my dream is to be able to work in Silicon Valley and learn from the best to go back to my country and teach them everything I know.
Physical inventory taking is a pain - even for our small store it takes us 4-5 people almost a full day to do that. Ideally I would love to walk through the store, record video and let it identify and count bottles. (Did dig deep into it and its a tough problem)
That’s a cool idea. That’d make it easier to do online shopping and local delivery, too. I would bet this is possible with some combination of cheap hardware, man power, and business sense. Identifying the ideal setup is the tricky/expensive part. Perhaps the problem can be broken down into smaller, more attackable pieces somehow. I’m no expert, by the way, just inspired by the optimism of the course.
Related to your idea, I wonder if anyone has tried making a crowd-sourced “Does the local store have this thing” service, where “buyers” can put a bounty on finding a certain product locally, and “finders” are people who are either present in the store, or willing to go there to check if the product is available. Kind of like mechanical turk for finding stuff.
Then you could work with a delivery service like Lyft/Uber, and boom, everyone’s knowledge about what’s where in any city can be made known to everyone. I feel this could help local stores at a time when people are more and more buying online. I’ll bet one could even give Pokemon points for accomplishing these real world tasks =). I also bet this already exists as some startup and I just don’t know how to Google it or it hasn’t broken through to world-wide popularity yet, which would be no surprise given the amount of new products coming out every day.
( I don’t think I have any truly original ideas, I just feel like this is a cool forum to brainstorm with like-minded folks who enjoy working with data )
Hi @sanjeev.b, I like your journey to DL and your ideas about DL application to retail. Walmart and its scanning robot confirm the big help of “intelligent video” in stores (big or not ).
TechCrunch article (10/27/17) : Walmart is rolling out shelf-scanning robots in stores, but says they won’t replace people
Yup true. Changes may come earlier in big retail because of pressure of Amazon and also they can afford to spend lot more on technology. But at the low end small retailer can benefit too - though the price point of the solutions has to be much lower (e.g. in the US a small retailer would be averse to spend more than say $100/month).