Full Transcript
Jeff Thompson:
Welcome to Blind Abilities. I’m Jeff Thompson. A few weeks ago, I was on the road, but I had my mobile recording studio with me. And I had an opportunity to do an interview with Envision co-founder, Karthik. Envision is the company that has developed Envision AI, which is a Swiss Army like tool for OCR object recognition, scene detection, it just does a lot of stuff.
Envision AI:
Describe scene. Dim. Looks like a green plant on a table.
Jeff Thompson:
It was great to sit down with Karthik and talk about the app, how it’s been developed, when it got developed, how Karthik & Karthik, yes, Karthik & Karthik brought this tool to the blindness community and wants your input on what they are doing. They thrive on feedback is basically what he’s saying. And following this interview, I give a demonstration of the Envision AI app, and try to highlight some of the main features, to give you a little bit of the flavor that Karthik & Karthik are bringing from Envision. And if you’re interested in more podcasts with the blindness perspective, check us out on the web at www.blindabilities dot com. Follow us on Twitter at blindabilities and download the free blindabilities app from the app store, that’s two words, blind abilities.
Jeff Thompson:
Enable the Blind Abilities skill on your Amazon device, just by saying enable blind abilities. So please welcome Karthik from Envision.
Jeff Thompson:
Welcome to Blind Abilities. I’m Jeff Thompson. In the studio with me is Karthik from Envision. How are you doing?
Karthik:
I’m doing good, Jeff. How are you doing?
Jeff Thompson:
I’m doing great. And thank you for taking the time to come on Blind Abilities and share with us about your app, Envision AI. Could you tell our listeners a little bit about the app?
Karthik:
Yeah. And firstly, thank you for inviting us. We’d love to go on different shows and just talk to the community in general. So Envision AI is an app that helps people with visual impairment to live more independently. So uses computer vision to help people read text, recognize faces, recognize objects and so much more. And it’s currently available free to download on IOS and Android as well.
Jeff Thompson:
Yeah, oh that’s great. Nice. I went through the list of languages. There’s a lot of languages that it does too.
Karthik:
Yeah. It’s 60 different languages. So the OCR is capable of reading 60 different languages. So whatever language you have on the iPhone, you can read it, and you can read all the languages that are there on Android as well. And that’s what we intend to do. To be like a global app, so if fits right in with what we wanted.
Jeff Thompson:
Well 60 languages, that pretty much covers the world right there.
Karthik:
Yeah. Exactly.
Jeff Thompson:
But it’s not just an OCR reader. Can you tell us some of the features that Envision AI has on it?
Karthik:
Yeah, sure. So Envision actually has two main features. One is with text recognition, so it’s text recognition of the OC part, where it’s able to read different kinds of text, like short pieces of text, or handwritten text, or like even books and documents. The other part is object recognition, so it’s able to detect different objects in your surroundings and then give you an idea about what it is. It can also describe scenes to you in actual languages, so it can say for example, a person looking at the camera, dogs sleeping on the floor. It has a bar code recognition feature, and it also has color recognition. So it does both as well. And we have more general recognition features coming to the app over the next few months.
Jeff Thompson:
Plus, when you were talking about scene recognition, that includes object and when I was looking at the app, I saw that you can actually tell it exactly what something is, if you want to identify and then submit that.
Karthik:
Yes. So you can actually go ahead and teach Envision faces of people. And when it tries to do a scene description, if it knows the person that is in the frame, it will go ahead and add that person’s name with the description itself. If, for example, like your face with Envision, and if I take a picture of you it probably says, looks like Jeff sitting in a room, blah blah blah.
Jeff Thompson:
Oh, that’s great. Another thing that I thought that really differentiates itself from other apps that I’ve seen out there is for low-vision people. The magnification, the color contrast, that’s really unique. And I was playing with that and even though you zoom in a little bit, it still keeps reading.
Karthik:
Yeah. It is. So again, like our focus was to always make this possible to use from everyone across the spectrum. We have also made the app capable of, even if you like increase the font size, do the maximum that you could do on an iPhone or an Android, it’s still capable of correctly adjusting the UI, it’s able to do DAC, it’s able, we have invert color as an option, so you can invert colors and be able to use the OCR on that. You can zoom, like you said, and do it so we focus not only on people who are like say users of Voiceover or Talkback, but we also make sure that the app is available for people with no vision as well. And going forward, we have a couple of other improvements coming to the low vision aspect of the app itself.
Jeff Thompson:
Oh really. That’s exciting. I was really impressed, because when I hit instant text, it started reading the keyboard on the laptop in front of me, then I held it up and it started reading the page that was on the website, so I brought it over to my mixer, and I kind of moved it around where I knew where some stuff was, and it was working really well.
Karthik:
Yeah. I mean that is actually, the instant text feature is one of the big I would say plus points with the app itself. And people use it in ways that we couldn’t even imagine when we put that feature out. Like people use it today to read subtitles off a screen. That’s how fast the feature is, right. We’ve had lots of users who just place the app a certain distance from the TV, and then it’s able to read subtitles out loud to them. They use it to read computer screens, because the OC is really good at reading different kinds of texts, is able to pick up computer screens as well. And people use that feature a lot, and it’s extremely empowering to a large number of users.
Jeff Thompson:
Yeah, I was impressed with that. Then you got the document section where it’ll actually help you locate the document.
Karthik:
Yeah. It does. So the document feature is again like something that a lot of people use heavily, especially people who work and have a lot of paperwork to read. So the app is capable of not just taking just single images. It’s able to take images in batch mode. So you can go ahead and scan multiple pages at a time, and then export all of them to any format you want and so on. The app is also capable of reading PDF, so you can import PDF almost any size to the app and you can start reading PDF in the app as well. And of course, all of this is open to 60 different languages, so no matter what language the document is in, it’s able to pick it up and read it out to you.
Jeff Thompson:
Oh that’s great. Especially for students transitioning to college and stuff. They can then use that for homework and stuff because batch mode, that’s what I would like to use that for, is when you got multiple, ten pages, just boom boom boom, get them all done, and then go back and read them. And like you said, you can export them to different formats. That’s awesome.
Karthik:
Yeah. This is actually really popular amongst students and professors and teachers. We’ve heard from a number of people saying that they use Envision on IOS and Android to read like one book a day. So there’s a professor that we spoke to recently from India, and he said that he uses Envision to read like massive three hundred, four hundred page PDF’s, and then make notes for his students with the help of the app, and he’s like, I just love using the app for this particular feature alone. And it’s really great truly at reading text.
Jeff Thompson:
PDF’s is important. Everyone’s always asking that question. Now there’s an answer for them, Envision AI.
Karthik:
Yeah. I mean even if your PDF has like images in the background or if, even if your PDF contains a lot of images, it’s still able to like correctly extract the text from the PDF, and then do the OCR and be able to give you a much better accuracy than most of the options out there.
Jeff Thompson:
Well that’s great. You mentioned when we first hooked up here, you had some future news coming.
Karthik:
Oh yeah. We’re really excited about some of the changes that we’re putting into object recognition within the app itself. So there is a tab called Scan and Find right now within the app, and we’re planning to really revamp that in the coming months. So Envision would not just be really good as reading text, but also really good at recognizing objects as well. And that’s our big aim for the app when it comes to IOS and Android in the coming months. And of course we’re going to constantly refine other areas of the app as well, but this is big. And also another thing that a lot of people would want to know is, when are we moving to the glasses aspect of it. So that’s a question that we get from a lot of users. And we’re also working on porting Envision to work with Smart Glasses, so that is also something that you can expect pretty soon from us.
Jeff Thompson:
Oh, that’s really cool. Then you’d just be looking with eyes.
Karthik:
Yeah.
Jeff Thompson:
Like you said, on your website it says, see what you can’t.
Karthik:
Yeah. And that’s what we aim to do with the glasses as well. So whatever you’re going to be able to do on the phone right now, you’re going to be able to do on the glasses as well. And we’re going to constantly keep updating the software on the glasses, so whatever new features come to the app, it’s also automatically going to be going to the glasses. So pretty exciting time for us here at Envision.
Jeff Thompson:
Oh that’s great. Well Karthik, is there anything else you’d like to share with our listeners?
Karthik:
I would really love to hear from your listeners about the experience using the app itself. We’re actually a company that’s very closely tied to the community. In fact, within the app, one of the most used features is this feature called request a call. So request a call allows you to go ahead and talk to us directly. And you can like request a call from us and we go ahead and give you call back as soon as we can. So this is a feature that both our users from the community and also we love, because it really connects the two of us.
Karthik:
I would like all of you guys to go ahead and try the app, and see if you like it, and we’d love to hear your feedback either through the request a call option, or you can just shoot us a feedback within the app also. So that’s my big ask to everyone in the community. Because we just love to hear from you guys.
Jeff Thompson:
Yeah. And they can try the free 14-day trial.
Karthik:
Yes. Envision is an app that’s free to install, so you can install it for free, and use it for 14 days and after 14 days, you can pick any one of our subscription plans, so we have plans starting from $5 per month to $40 per annum, and $200 if you just want to buy it out for a lifetime.
Karthik:
We constantly keep putting out updates almost every two weeks now. So yeah, you guys, if you have any questions or concerns, we’re always on top of it, and we always make sure that the app is as bug-free as possible.
Jeff Thompson:
Oh Karthik, I was using this earlier today, and I was moving around the room and doing some testing it out, giving it some, it’s crisp. What I mean by that is, what it reads, it just reads it, it’s got good voice, I understood it right away. The learning curve was next to nothing. I just started using it and it’s pretty self-explanatory. And I really like that you could program something in it to identify something. And I think in the future like you said, find something. That’d be really cool, just go around and find it. So good job.
Karthik:
Thanks. So that’s been only possible because the community has been constantly giving us feedback. And that’s how we’ve been really improving because each person who uses Envision and tells us about a feature that they would like to see, or something that they would like to see improved, we have always been doing that. That’s the reason why I feel today like, you know, when you say the app is crisp, just because of the community.
Jeff Thompson:
What got you started on developing an app that would do all this?
Karthik:
Envision as an idea started way back in India in 2016, in late 2016. When me and my co-founder, we went to a school for the blind in my hometown back in India, and my co-founder was giving a talk there about being a designer, and how being a designer means being able to solve problems. And designing is just problem-solving. And at the end of the talk, when he asked the kids there as to what kind of problems they’d like to solve, I think a lot of them spoke about being able to read books independently, being able to go out with their friends independently, and just having the option to pursue whatever career they wanted. And that got us thinking a lot about independence and that’s when we started to put together our ideas on what’s happening in AI and his ideas on what’s happening in design. And the two of us came together to start working on it.
Karthik:
At that time, my co-founder, his name is also Karthik, he was studying industrial design in Delft, in the Netherlands. And he came back to Delft after this vacation in India, and he started working on Envision as a masters thesis, and eventually, the two of us started working together on early prototypes of Envision, and towards the end of his thesis, when we started showing Envision to a lot of people, they just loved it. And they said, if you guys really want to see this make a difference, you guys should start this as a company, you guys should do this full-time. And that’s when we started Envision in October of 2017. And here we are, with the help of the community.
Jeff Thompson:
Oh I think it’s a great initiative. There’s so many people there that just want something to access the same stuff everybody else accesses.
Karthik:
Exactly.
Jeff Thompson:
And I could tell right away, when I fired up the Envision AI app, it was working. I didn’t have to fumble around, so anybody out there that’s listening, give it a shot. Do the trial, consider it. I think it works very good. I think it works better than most of them out there.
Karthik:
Yeah. I think our aim at Envision, our long-term vision is to make all visual information as accessible as possible. That’s what we’re working towards every day. And what we’re doing right now fits perfectly well with that vision. We’re pretty excited to come to work each day and we see a lot of the impact, and that really drives us to work even harder on the app, and you can expect a lot more going forward.
Jeff Thompson:
Well great. We’re all looking forward to it. Perfect. I want to thank you very much for you, and the other Karthik too, for what you guys are doing for the visually impaired and blind community, because this app really works well. I’m glad to have it on the show and talk about it here. And thanks again for coming on and talking about Envision AI. And I also like that it’s available for Android and IOS, so great job you guys.
Karthik:
Yeah, thank you so much, Jeff. And thank you for reaching out to us, and you know, we’ve been listening to Blind Abilities also, so we cover, we listen to a lot of podcasts that we find interesting within the community to help us stay connected and Blind Abilities is one of them, and so definitely looking forward to coming back here again with a lot more news.
Jeff Thompson:
Well we will surely have you on. All right. Thank you very much.
Karthik:
Yeah, thank you so much, Jeff. Have a great day.
Jeff Thompson:
So let’s take a look at Envision AI. Now this is available on Android and IOS. And it comes with a 14 day trial. So download it, put it through its rigors, see how it works for you. There’s multiple ways of subscribing. You can subscribe by one month, for $4.99 a month, or they now have a six month for $29.99, and annual for $39.99, and a lifetime for 199 dollars. So let’s take a look at it.
Envision AI:
Envision.
Jeff Thompson:
When you open it up, it starts up with magnification up in the upper left-hand corner.
Envision AI:
Magnifier.
Jeff Thompson:
Tap on that, and whatever you’re looking at magnifies. And you adjust the magnification, now here’s pretty good. Pretty steady. And you can adjust the contrast. That’s very good for low vision users. So let’s go down to the bottom. Now there’s going to be four sections here. The first one’s going to be text, the next one is general, the next one is scan and find, and then help. I got text selected right now. So let’s do a single finger double tap.
Jeff Thompson:
From left to right near the bottom, you’re going to find start reading instantly. And that’s where you just pass it over some text, and when it’s detected, it’ll start reading to you.
Envision AI:
Start reading instantly. Button. Start reading instantly. Curb sideways P.O. Box 43000, 154 Brooklyn Park, Minnesota 55443-0154.
Jeff Thompson:
All right. Now I know that a piece of mail that has to go out. Then we have-
Envision AI:
Read handwritten text. Button.
Jeff Thompson:
You know, stuff like greeting cards and stuff. That comes in handy. Then finally you have-
Envision AI:
Read document. Button. Read document. Take photo. Bottom and left edge is not visible. Bottom and left edge is not visible. Processing. A Different Way of Seeing, a blind woman’s journey of living an ordinary life in an extraordinary way, Lois Strachan.
Jeff Thompson:
But let’s go back to read document button one more time.
Envision AI:
Read document. Button.
Jeff Thompson:
But if you wait for it-
Envision AI:
Double tap and hold to read multiple pages.
Jeff Thompson:
And that’s the one I’m going to demonstrate right now.
Envision AI:
Would you like to read multiple pages? Yes. Button.
Envision AI 2:
All edges visible.
Envision AI:
Take photo.
Envision AI 2:
Hold steady. All edges visible. Hold steady. Bottom and right edges not visible. All edges visible. Hold steady. All edges visible. Hold steady.
Envision AI:
Done taking photos. Button.
Envision AI 2:
All edge-
Envision AI:
Done taking. Processing.
Envision AI:
Endorsed by the South African National Council for the Blind, and the South African Guide Dogs Association for the Blind. Have you ever wondered how a blind person pours a cup of coffee? Or how they and their guide dog know when it’s safe to cross a busy road? When Lois Strachan lost her sight at the age of 21 years, she had to learn the answers to these and countless other questions of how to live as a blind person in a sighted world. In this delightfully quirky and entertaining book, Lois shares some of the secrets she discovered about how to live an ordinary life in an extraordinary way despite her blindness. An entertaining, informative and educational read that gives you an insight into the life of an inspirational woman who lives life to the fullest. Well worth reading.
Jeff Thompson:
It seems pretty quick. Maybe a little too quick. As you want to turn the pages, once you do capture it all, you are able to go from page to page.
Envision AI:
Next page. Button. Previous page. Button.
Jeff Thompson:
This may be handy, especially when you have multiple pages that you want to scan in. You can share it to multiple places on your phone or elsewhere.
Envision AI:
Export text. Button. Message. What’s App. Mail. Pocket. Add to Notes. Twitter. Facebook. One note.
Jeff Thompson:
I also had great success with the PDF. I went into my photos, found a PDF and I opened it with the Envision AI app. And it worked perfectly. Color detection has always been iffy. They give you a choice here, from a basic 30 colors, kind of like a box of crayons, up to 900 different color variations.
Envision AI:
Detect colors. Button. Detect dark. Medium gray. Gray. Warm gray. Purplish gray. Slate gray.
Jeff Thompson:
Quite a spectrum.
Envision AI:
Sand brown. Pale brown. Dust. Very light brown. Taupe. Gray brown. Warm brown. Milk chocolate. Chocolate brown. Brown.
Jeff Thompson:
You can detect and read bar codes.
Envision AI:
Scan bar code. Button. Scan bar code. The result, Dura lock Power Preserve. Describe scene. Button.
Jeff Thompson:
To test this, I took a picture over towards my window, where I know there’s a plant.
Envision AI:
Describe scene. Dimmed. Looks like a green plant on a table. Scan and find. Tab. Recognize common objects.
Jeff Thompson:
To test this, I decided to just scan around the room by turning in my chair.
Envision AI:
Room. Desk. Musical instrument. Flower pot. Drawer.
Jeff Thompson:
Okay.
Envision AI:
Teach Envision. Button. What would you like to teach? Teach a face. Button. Zero faces. Zero faces. Switch to front camera.
Jeff Thompson:
I switched cameras, and took multiple pictures of myself.
Envision AI:
One face in frame near one foot away. Take photo. Button. Take photo. Add four more images. Take photo. Button. One face in frame near one foot away. Take photo. Take photo. Take photo. Take photo. Done taking photos. Enter name to take button.
Jeff Thompson:
Jeff Thompson.
Envision AI:
Inserted Jeff Thompson. Alert. Teaching successful.
Jeff Thompson:
Since I’m not used to the built-in voice, I decided to go into the help section and change the voice to something I’m more familiar with.
Envision AI:
Speech. Button. Speaking rate. Fifty percent. Adjustable. Voices. Button. English. Button. Voices. Aaron. Selected. Alex, Allison, Allison enhanced, Arthur, Ava, Ava enhanced, Catherine, Daniel, Fred, Gordon, Karen, Martha, Moira, Nicky, Nicky enhanced, Samantha, Samantha enhanced.
Jeff Thompson:
And then I decided to see if it would find me amongst the room.
Envision AI:
Recognize custom objects.
Envision AI 2:
Jeff Thompson.
Jeff Thompson:
Okay. I am a custom object. Now let’s see if I am a common object.
Envision AI:
Recognize common objects.
Envision AI 2:
Musical instrument. Room. Selfie. Jeff Thompson. Jeff Thompson. Cool.
Jeff Thompson:
I guess it is a selfie. But did you hear that? I’m cool.
Envision AI 2:
Jeff Thompson. Cool.
Jeff Thompson:
One more time.
Envision AI 2:
Jeff Thompson. Cool.
Jeff Thompson:
Hey. My overall feeling of Envision AI is it’s pretty snappy. Maybe too snappy. I would like to see a few features tweaked a little bit, such as in batch mode, when you’re going to take multiple pictures, I would like to have more control when I’m ready to do it. It seemed to be taking pictures pretty quick and I wasn’t that quick. I’m going to give them some feedback on the whole entire thing. I know that object recognition is kind of basic. That’s across the board with most apps, but they say they’re going to be improving this, so I’m looking forward to seeing that. Overall, I think it’s a pretty good app. I think it’s one of the better OCR apps out there, as for reading, detection of the print, and usability. So good job Envision AI. Remember, there’s a 14 day trial. Use that 14 day trial and decide if this is the app for you.
Jeff Thompson:
A big shout out to Chee Chau for his beautiful music. You follow Chee Chau on Twitter @lcheechau Chee Chau, Chee Chau.
[Music] [Transition noise] -When we share
-What we see
-Through each other’s eyes…
[Multiple voices overlapping, in unison, to form a single sentence]
…We can then begin to bridge the gap between the limited expectations, and the realities of Blind Abilities.
Jeff Thompson:
For more podcasts with the blindness perspective:
Check us out on the web at www.BlindAbilities.com On Twitter @BlindAbilities
Download our app from the App store:
‘Blind Abilities’; that’s two words.
Or send us an e-mail at:
Thanks for listening.
http://blindabilities.com/category/stem/
Contact:
Thank you for listening!
You can follow us on Twitter @BlindAbilities
On the web at www.BlindAbilities.com
Send us an email
Get the Free Blind Abilities App on the App Store.