Podcast: Play in new window | Download
Your weekly dose of information that keeps you up to date on the latest developments in the field of technology designed to assist people with disabilities and special needs.
Show notes:
Alberto Rizzoli Co-Founder Aipoly | www.aipoly.com
7 more terrific Android apps you can’t get on the iPhone http://buff.ly/1Qo9Mph
The mobile phone of the future will be implanted in your head – CNET http://buff.ly/1JysHy6
App: Uber – www.BridgingApps.org
——————————
Listen 24/7 at www.AssistiveTechnologyRadio.com
If you have an AT question, leave us a voice mail at: 317-721-7124 or email tech@eastersealscrossroads.org
Check out our web site: https://www.eastersealstech.com
Follow us on Twitter: @INDATAproject
Like us on Facebook: www.Facebook.com/INDATA
——-transcript follows ——
ALBERTO RIZZOLI: Hi, this is Alberto Rizzoli, and I’m the cofounder of Aipoly, and this is your assistance technology update.
WADE WINGLER: Hi, this is Wade Wingler with the INDATA Project at Easter Seals crossroads in Indiana with your Assistive Technology Update, a weekly dose of information that keeps you up-to-date on the latest developments in the field of technology designed to assist people with disabilities and special needs.
Welcome to episode number 244 of assistive technology update. It’s scheduled to be released on January 19 of 2016.
Today I talk with Alberto Rizzoli who is the cofounder of Aipoly, an object recognition app that I think you will really enjoy. We’ve got a story about android apps that you can’t get on the iPhone; and the folks over at BridgingApps talk with us a little bit about Uber.
We hope you’ll check out our website at www.eastersealstech.com, sent us a note on Twitter at INDATA Project, or give us a call on our listener line. We love to hear you. That number is 317-721-7124.
If you’ve listened to the show before, you know that I am at Apple and iOS fan boy, so it hurts me but here is an article called, “Seven more terrific android apps you can’t get on the iPhone.” It comes from BGR Technologies, and we found out about it from our friends over at top tech tidbits. Basically it runs through a list of apps that you can’t get on the iPhone but they feel like you should. They are on android. One is called twilight that allows you to adjust the luminescence of your screen as the day goes on; another called anticipate that loads webpages faster for you; a service called hound beta which is supposed to be a voice-enabled personal assistant that puts Siri and Cortana to shame; one called Pender that aggregates all of your food delivery services into a simple app that apparently learns more about what you eat and how to get to you have faster; one called Greenify that adjusts apps that are taking up too much battery usage and dropping them down in the background; and another called dog catcher which I think is interesting because it’s a podcatcher for your android phone so you can put assistive technology update and your other favorite assistive technology podcast right there in dogcatcher. A couple more on the list I didn’t get to, but I will pop a link in the show notes over to them so if you’re on android you’re going to enjoy these apps. If you’re on iOS, sorry, not there.
So falling into the category of kind of creepy but kind of cool is a headline from CNET.com that says, “The mobile phone of the future will be implanted in your head.” Apparently within the last week or two, there was a group getting together at the world economic forum, and there were a bunch of future thinkers and industry leaders who basically said that the next logical step with technology innovation is to have implantable cell phones. It’s not totally out of the ordinary. We have cochlear implants, pacemakers, implantable glucose monitors to help people with diabetes monitor their blood sugar. There are some folks who agree that cell phones or at least wireless technology that allows for communication will be implanted sometime in the next several years. Eric Brynjolfsson from MIT says that computers and other digital advances are doing for mental power, the ability to use our brains to understand and should our environments, what the steam engine and its descendants did for muscle power. There are predictions that by 2026, one in every 10 cars on the US roads will be driverless and that’s just a decade from now. This group made a lot of other predictions or responded to servers that tell us that technology is moving along at a pretty rough clip. I’m going to put a link in the show notes over to the CNET article where you can read more about what they are saying about mobile phones in the future might being implanted in your head. Check our show notes.
Each week, one of our partners tells us what’s happening in the ever-changing world of apps, so here’s an app worth mentioning.
AMY BARRY: This is Amy Barry with BridgingApps, and this is an app worth mentioning. Today I am sharing Uber. Uber is an app for people who are looking for an alternative to a car service or city taxi cab. The android, iOS, and Windows phone app connects riders with drivers using their phones GPS capabilities, letting both parties know one another’s location and removing the question of when the ride will actually arrive. In addition, the tech company also processes all the payments that are involved, charging the passengers credit card, taking a cut for itself, which typically ranges from 5 to 20 percent, and then direct deposit in the remaining money into the drivers account. It’s all in the background and completely cashless. Depending on availability, Uber also offers several different levels of service. The Uber service is for anyone who uses public transportation, and have found it can be especially great for older adults and individuals with physical disabilities or special needs who may not be able to drive or live in a city where public transportation is a struggle. For riders with physical disabilities, labor may be their primary means of transportation, and it provides independence they may not have otherwise had. Uber recently rolled out product innovations that facilitate greater economic opportunity for deaf and hard of hearing drivers. Visible and vibrating alerts aid deaf and hard of hearing riders. In-app features such as the ability to enter a destination ensure effective communication between the rider and the driver. The Uber rider’s app’s accessible technology features accommodate people who are blind and visually impaired, and they are working with their driver partners to ensure service animals are always welcome. With voiceover iOS, and wireless braille display compatibility, the Uber app provides a really reliable transportation option for the blind and visually impaired community. Uber Assist is also available in select markets and is designed for riders who would like additional assistance. All of these drivers accommodate for the wheelchairs. Driver partners receive specialized training from disability community organizations to provide additional assistance for seniors and people with disabilities. Uber is currently piloting Uber Assist in Chicago, Houston, Diego, Honolulu, Los Angeles, San Francisco, and Portland Oregon, and the rides are the exact same cost as regular Uber X services. Our reviewer has been using Uber with older adults who live in New York City. They really like the cleanliness of the Uber cars, and knowing the amount of money that they will be paying for the fair. They find that the app is super easy to use, most of the time. Occasionally they do find that it may be difficult during rush hour or implement whether. But overall, the app has been a go to app, especially in the colder weather, or when they are traveling a long distance. BridgingApps highly recommends the Uber app. Uber is a free download on android, iTunes, and also the Windows phone, and then the cost of the service varies. For more information on this app and others like it, visit BridgingApps.org.
WADE WINGLER: What you are about to hear is just a quick sampling of the voiceover reading of the Aipoly app running on my iPhone. We are getting ready to talk to the creator of the Aipoly app, but before we did that I thought it would be cool for me to just turn on the app, turn on voiceover, and then point it at different things in our recording studio to see if it would recognize the objects, because we’re going to learn a ton about object recognition. Before we jump into the interview, listen to what this app says.
VOICEOVER: Thermostat. MacBook Pro. CRT. AC. Trashbin. I’m not sure. Pen. Apple Mac. Keyboard. Boots. Walls. Hand. Index. Face.
WADE WINGLER: Object recognition is something that we had been paying attention to for a while here and obviously our listening audience has been fascinated by some of the conversations surrounding object recognition. Recently I downloaded an app called Aipoly, turned it on my iPhone, and was quite frankly astounded at how it was able to recognize objects in my area. In fact, leading into this interview, you heard some of the Aipoly app talking about some of the objects that were recognized in my studio here, everything from my foot to a trashcan and some computers and some things as well. I am so excited to have Alberto Rizzoli on the phone today, on Skype or the Internet, some platform we’re using today. He is the cofounder of Aipoly. He is joining me, stayed up late and I got early, from Melbourne Australia right now. We are so excited to have them on the phone. Alberto, welcome to the show.
ALBERTO RIZZOLI: Hi, thank you so much.
WADE WINGLER: I have been messing around with Aipoly, and it’s pretty cool. We recently have had a couple of conversations on our show about object recognition. Obviously something that is useful for people who are blind or visually impaired, there are other opportunities for folks to benefit from that technology. But I want to know first about you and talk to us about Aipoly. Tell me a little bit about you and your background and how you came to the place where we are talking about this today.
ALBERTO RIZZOLI: The reason why we started Aipoly is part of a mission that we underwent at singularity University in Silicon Valley, the one impacting the lives of 1 billion people in a positive way within 10 years. Every student in this institution is selected from across the world. I represented Italy. My cofounder represented Australia and Sweden. We got together at the NASA Research Park where they once tested the stuff planes to develop artificial intelligences. While we were looking into how they could actually see today, we thought well, how can we use this brand-new technology to have the most positive impact on people’s lives? I was coming from a background in which we all interact in one way or the other with people who are blind or visually impaired. We really thought this could improve people who cannot see. In only a few years, this can be revolutionary.
So we started Aipoly. We built a prototype. It was a huge success. This was August of last year [2015], and now six months later we have an app on the app store and a whole range of very exciting updates coming down the line.
WADE WINGLER: So my initial experience has been so far, I downloaded it, turned it on, it was very simple to understand. Then I started pointing it at things to see what it would recognize, everything from objects in my house to Mickey Mouse and those kinds of things. So my experience is bent I pointed out things and it tells me fairly often what those items are with a pretty good level of detail. Tell us a little bit about how it works. Was going on there?
ALBERTO RIZZOLI: Behind Aipoly, there is something called a convolutional neural network. I know it’s a very long word, but have no fear. They will become more popular as time passes. They started becoming more popular around 2011 when some researchers mostly in California and realized this process actually looks like a simulation of the human brain, and when it’s done by computer, it can do some very incredible thought processes. This can include anything from detecting the mood that is in a written paragraph on the way to detecting something within a photograph and an image. So this neural network, this tiny brain, so to speak, inside Aipoly has been trained, much like a baby. It’s seen a huge number of in its infancy, and that is what it’s able to remember and approximate. It has been 10 million different images from a very carefully selected data set that we and our partners put together, so it has a very good knowledge of the world, especially the indoor world. When Aipoly sees something that I can remember in some way, it uses a variety of different statistical methods to think, that actually looks like one of those 10 million images that I’ve seen. That chair looks familiar. Then item looks familiar. It could be perhaps a bicycle. These probabilities improve over time. But at the moment, it can tell about 1,000 different things apart with pretty good accuracy. If you compare this to an adult, which is the primary user of Aipoly, there are so many more things that we know. That is why we need to help artificial intelligence in learning more object for time.
WADE WINGLER: That makes sense. Tell me a few examples of how people are using the Aipoly tool right now, some common uses and maybe some uncommon examples.
ALBERTO RIZZOLI: Some common uses are the inbuilt color detector can be used directed by different shapes. Color detectors have been around for a while, but using it in combination with object detection, people can actually look for a different item and recognize the color. One example is we were experimenting with different types of identifiers here at Aipoly, which we haven’t released. Just imagine if you are able to have an AI that specifically identifies fruit, you can actually pick up a grapefruit and you’re able to tell if it’s a pink one or a yellow or white one. You’re able to tell a lemon from a lime, which is something that requires both shape and color.
Some unusual cases for Aipoly is that people seem to really enjoy the fun know that we initially created to demo. There is a bit of a personality behind it, but a lot of people, especially sighted users, and especially children, have really enjoyed it. We actually rolled out a brand-new update to the app that is going to be released on Monday that adds more responses to the fun mode, as well as many other new features that our users have suggested to us through feedback.
WADE WINGLER: And I have to admit that I downloaded the app and was playing with it last night before coming in for this interview. I turned on fun mode and get a big kick out of it when it did two things. I have a son who is four years old who has really big curly hair, and it recognized him from the background as a tabby cat. It’s that someone like why are you showing me a tabby cat or something like that. We got a big kick out of it when it misrecognized that. Then he turned around in its interface and it knew what he was. But we got a big kick out of why are you showing a tabby cat. As we were messing around with this app at the house last night, I simply didn’t – I handed it to my wife. I didn’t tell her what it was, what it was for, didn’t say anything except just turned it on and it started recognizing things. She wandered around for a while playing with it and having a ball seeing what it could recognize and what it couldn’t. She had a couple questions. One, what is the weirdest place that you have used Aipoly. And then she also wanted to know, does it connect landmarks like the Eiffel Tower or the Statue of Liberty.
ALBERTO RIZZOLI: Great question. The weirdest place where we used Aipoly is in Las Vegas. We were just invited by the Consumer Electronics Association for being one of the top startups in helping people with a disability. We were invited there, and this was the largest lecture next show in the world. It was very different from — light all over the place and objects that in no way possible could fit inside the data set. There are technologies that we’ve never seen before. Simply the layout of Eureka Park where we are located, it looks like nothing else. It doesn’t look like at home, it doesn’t look like an office. It is simply outlandish. Imagine now putting a two-year-old or three-year-old, which is about the current intelligence of Aipoly, in the middle of this place — even though it has this hunky British accent at the moment, it still does have the brain of a very young person — and then telling it recognize things for me. Tell me what is around. So we have seen the app get very confused while trying to identify the various technology. It still didn’t work. Even though there were about 16 different points of light at the same time rather than the usual three or four that you have in a normal setting — light is a very important element actually for object recognition. If you have too many shadows in the wrong places, the computer gets confused. It was very string cheese at Consumer Electronic Show, even though it performed quite well. We spent a lot of time trying to remove false positives, the biggest negative when using these object recognition apps, especially those in real time Aipoly, which is a brand-new innovation. We never want it to say that someone is a tabby cat. But this can happen.
Aipoly is only a couple of weeks old. We’ve only developed it for the past few months, and the technology is brand-new. But over time it will stop seeing tabby cats. The baby within it will grow. We are very excited about that.
WADE WINGLER: I have to agree. I have spent a lot of time in Vegas. I worked out there often on for about 10 years. There are many things walking around that I wasn’t able to recognize either. I’m not going to hold Aipoly to too high of a standard. What about the landmarks, things like the Eiffel Tower or Statue of Liberty? Is it there yet?
ALBERTO RIZZOLI: Regarding landmarks, at the moment Aipoly can’t recognize most of them. We have a couple of landmarks in the data set, but I’ll keep them as an Easter egg. If someone ever finds and points their phone at a landmark and suddenly it says what it is, then you found it, you win the game. I’ll keep those secret. However, we are experimenting with a cloud-based service that recognizing landmarks. What is special about Aipoly is it does not use the Internet. The entire intelligence is in your phone. We are looking into what the people want to recognize landmarks from afar if they can rely instead on an online tool. This can help not just with landmarks, but imagine the future, if you’re an art enthusiast, whether you are sighted or not, you can actually identify things within a museum by simply using your phone and computer vision. Landmarks can’t be notified today, but they are a possibility in the future and we have done a couple of experience on it. They are a lot of fun.
WADE WINGLER: It sounds like the accuracy is going to get better. Another funding that we did last night at home was we pointed it at a mini mouse statue, just a little toy, and at first it said chandelier and then it said Mickey. Mickey was very close to what we were looking for. Chandelier, then I thought why would you think that? And I thought the big round orbs and the shape of a Mickey Mouse head might kind of look like a chandelier. When it misrecognized, I could at least understand what it might ink that way. Talk to me a little bit about accuracy and how accuracy is getting better.
ALBERTO RIZZOLI: Accuracy is getting better, and parts thanks to our users, because if you are using Aipoly and you are sighted, then you can actually teach it by identifying any object. If it misidentifies something or it says I’m not sure, which is the response that it prompts one it is actually identify something but the probability is too low, then you can press a button and you can add a caption and you can at this to the entire knowledge of Aipoly, so that everyone in the world can benefit from your contribution. We have had users in Germany who have added hundreds of objects in their house. We have had engineers. We had one engineer in Austria who picked up all of their tools and meticulously named every single one of them. We are getting some great entries. If there is enough knowledge from around the world, each one of these will actually be remembered by the system. It will get smarter over time. We are predicting that between this and the work of ourselves and our partners in making a better dataset, bigger datasets, and increasing the number of things that can identify. We are looking into Aipoly and define not just an incremental 10 percent, but something in the thousands of percents of more objects than the coming months. We are very excited about that.
WADE WINGLER: Wow. Alberto, we are getting a little close on time for the interview here, but I have a couple of questions that I just had to ask. First of all, some practical questions: what platforms is it available on, and what is the cost? What the current availability like?
ALBERTO RIZZOLI: Aipoly is absolutely free and will remain free for iOS. We are working on an android version, and it’s going really well and really fast. That will also be available for free in the coming weeks.
WADE WINGLER: Excellent. As you look out into the future for Aipoly and the possibility of things they might do, what is in your crystal ball? What do you see coming down the pike for Aipoly?
ALBERTO RIZZOLI: For example, one thing that we built a test for is, what if you could learn a language by simply waving your phone in front of objects. The synonym of an object in Japanese, and you could read the transition in real time and quickly learn positive terms in a different language using your visual memory if you are sighted. We are also helping people who cannot remember the names of objects after they have had a stroke or accident. We are also looking into early childhood development and potentially learning disabilities to teach the name of objects.
WADE WINGLER: I think that is fascinating. As I was playing around with it, I was wondering, one of these days it should have OCR in it as well so I can only tell what it is, but I can tell it’s a can of tomato soup or something like that. That’s my feature request.
ALBERTO RIZZOLI: Stay tuned because it is something that we have worked on. Yes, indeed. Recognizing a can and telling what it is, is definitely one of the most impressive features, and he listened and are working on it.
WADE WINGLER: That is great. Alberto Rizzoli is the cofounder of Aipoly and has been our guest today. Alberto, how can folks learn more? If they want to follow your story, want to download the app, do you have a website or other contact information you would like to provide?
ALBERTO RIZZOLI: Absolutely. Go on Aipoly.com. You can download the app for free from there. You can follow us on Facebook. We post regular updates on what we are working on. It’s going to be a lot of fun to follow what Aipoly will learn next.
WADE WINGLER: Thank you so much for staying up late and being with us today. We appreciate your being on the show.
ALBERTO RIZZOLI: Thank you, Wade. It was a pleasure.
WADE WINGLER: Do you have a question about assistive technology? Do you have a suggestion for someone we should interview on Assistive Technology Update? Call our listener line at 317-721-7124. Looking for show notes from today’s show? Head on over to EasterSealstech.com. Shoot us a note on Twitter @INDATAProject, or check us out on Facebook. That was your Assistance Technology Update. I’m Wade Wingler with the INDATA Project at Easter Seals Crossroads in Indiana.