[Script Info]
Title: 
[Events]
Format: Layer, Start, End, Style, Name, MarginL, MarginR, MarginV, Effect, Text
Dialogue: 0,0:00:00.00,0:00:05.52,Default,,0000,0000,0000,,Module 10.3, examples of start-ups that \Nuse signal processing as a core
Dialogue: 0,0:00:05.52,0:00:09.50,Default,,0000,0000,0000,,technology. \NEarlier on in this class somebody asked
Dialogue: 0,0:00:09.50,0:00:13.01,Default,,0000,0000,0000,,in the forum, if I follow digital signal \Nprocessing class, can I get a job in the
Dialogue: 0,0:00:13.01,0:00:18.28,Default,,0000,0000,0000,,start-up and what sort of start-ups? \NSo this brought us to think, well maybe
Dialogue: 0,0:00:18.28,0:00:23.55,Default,,0000,0000,0000,,we can describe a few start-ups that came \Nout of research done in the lab.
Dialogue: 0,0:00:23.55,0:00:28.23,Default,,0000,0000,0000,,And our four that we discuss here, there \Nare actually more that are active, but
Dialogue: 0,0:00:28.23,0:00:32.41,Default,,0000,0000,0000,,four will be discussed here are \NIllusonic, Quividi, Sensorscope and
Dialogue: 0,0:00:32.41,0:00:36.62,Default,,0000,0000,0000,,Vidinoti. \NSo the first start up I want to discuss
Dialogue: 0,0:00:36.62,0:00:40.44,Default,,0000,0000,0000,,is called Illusonic. \NIt was started by Cristof Faller who did
Dialogue: 0,0:00:40.44,0:00:43.82,Default,,0000,0000,0000,,his PhD thesis on a time as a \N[INAUDIBLE], and was interested in
Dialogue: 0,0:00:43.82,0:00:51.06,Default,,0000,0000,0000,,acoustical signal processing. \NAnd in particular in multi channel audio.
Dialogue: 0,0:00:51.06,0:00:54.02,Default,,0000,0000,0000,,So if you do acoustical signal \Nprocessing, there are questions like
Dialogue: 0,0:00:54.02,0:00:57.67,Default,,0000,0000,0000,,beamforming, echo control, we just \Ndiscussed this earlier.
Dialogue: 0,0:00:57.67,0:00:59.46,Default,,0000,0000,0000,,We used a project of can you hear, the \Nshape of a room.
Dialogue: 0,0:00:59.46,0:01:03.66,Default,,0000,0000,0000,,When you want to do spatial audio \Nprocessing, you want to generate audio
Dialogue: 0,0:01:03.66,0:01:08.94,Default,,0000,0000,0000,,for many channels. \NEither for headphones or for multichannel
Dialogue: 0,0:01:08.94,0:01:14.02,Default,,0000,0000,0000,,loudspeaker systems, you may want to do \Nupmix or you take a stereo signal and you
Dialogue: 0,0:01:14.02,0:01:20.44,Default,,0000,0000,0000,,would like to render it as a 5:1 signal \Nor as a 17:1 signal.
Dialogue: 0,0:01:20.44,0:01:25.84,Default,,0000,0000,0000,,And, there are tools, of course, where, \Nyou can use signal processing techniques,
Dialogue: 0,0:01:25.84,0:01:33.07,Default,,0000,0000,0000,,for example, to de-noise Music or \Nde-reverb, recording of, person singing.
Dialogue: 0,0:01:34.11,0:01:38.34,Default,,0000,0000,0000,,So the tools that are used at Illusonic \Nare classic digital signal processing
Dialogue: 0,0:01:38.34,0:01:42.82,Default,,0000,0000,0000,,tools, plus what was discussed briefly, \Nwhen I talk about the class on audio and
Dialogue: 0,0:01:42.82,0:01:49.14,Default,,0000,0000,0000,,acoustic signal processing. \NAgain, perceptual models are extremely
Dialogue: 0,0:01:49.14,0:01:52.85,Default,,0000,0000,0000,,important because the human auditory \Nsystem is a very sophisticated signal
Dialogue: 0,0:01:52.85,0:01:56.86,Default,,0000,0000,0000,,processing device, and if you try to fool \Nthat device you better need to understand
Dialogue: 0,0:01:56.86,0:02:04.32,Default,,0000,0000,0000,,how it works. \NHere is an example of cool application,
Dialogue: 0,0:02:04.32,0:02:09.38,Default,,0000,0000,0000,,so let's say you have your home cinema \Nand you have a stereo recording that you
Dialogue: 0,0:02:09.38,0:02:16.41,Default,,0000,0000,0000,,would like to listen to. \NSo the home cinema has actually in this
Dialogue: 0,0:02:16.41,0:02:21.44,Default,,0000,0000,0000,,case one, two, three, four, five, six, \Nseven, eight, nine plus probably two base
Dialogue: 0,0:02:21.44,0:02:26.31,Default,,0000,0000,0000,,booster somewhere, so it's probably an \Neleven channel system, so you would do
Dialogue: 0,0:02:26.31,0:02:34.46,Default,,0000,0000,0000,,enough mix from a stereo signal, let's \Nsay from your MP3 player...
Dialogue: 0,0:02:34.46,0:02:38.05,Default,,0000,0000,0000,,To this eleven channel spatial audio \Nsystem, and you would like to make it so
Dialogue: 0,0:02:38.05,0:02:42.00,Default,,0000,0000,0000,,that it sounds really like you're in the \Nconcert hall.
Dialogue: 0,0:02:42.00,0:02:44.91,Default,,0000,0000,0000,,And so even sony has a very cool \Ntechnology to do this, and not only do
Dialogue: 0,0:02:44.91,0:02:47.93,Default,,0000,0000,0000,,they have the technology, they actually \Nsell a box that will do this at
Dialogue: 0,0:02:47.93,0:02:55.18,Default,,0000,0000,0000,,professional quality level. \NSo the company is it's a small company
Dialogue: 0,0:02:55.18,0:02:59.40,Default,,0000,0000,0000,,about five people, half a dozen people, \Nit licenses technology, state of the art
Dialogue: 0,0:02:59.40,0:03:05.78,Default,,0000,0000,0000,,stuff, to other, companies, and it has \Ncustom technologies that it develops.
Dialogue: 0,0:03:05.78,0:03:09.94,Default,,0000,0000,0000,,For specific applications and as I \Nmentioned it has this very cool Immersive
Dialogue: 0,0:03:09.94,0:03:14.16,Default,,0000,0000,0000,,Audio Processor that was just launched \Nthis year and please visit our website
Dialogue: 0,0:03:14.16,0:03:18.06,Default,,0000,0000,0000,,and see this cool stuff and if you want \Nto buy one of these Immersive Audio
Dialogue: 0,0:03:18.06,0:03:24.66,Default,,0000,0000,0000,,Processors, I can tell you it sounds \Nincredibly beautiful.
Dialogue: 0,0:03:27.01,0:03:29.05,Default,,0000,0000,0000,,The next company I want to describe is \NQuividi.
Dialogue: 0,0:03:29.05,0:03:32.63,Default,,0000,0000,0000,,Now this is a very important company in \Nits class because its a company of Palo
Dialogue: 0,0:03:32.63,0:03:36.79,Default,,0000,0000,0000,,Prandoni. \NSo when he's not teaching on Coursera and
Dialogue: 0,0:03:36.79,0:03:41.53,Default,,0000,0000,0000,,playing his his guitar To explain signal \Nprocessing.
Dialogue: 0,0:03:41.53,0:03:46.17,Default,,0000,0000,0000,,He's actually the CTO of a company in \NParis, called Quividi.
Dialogue: 0,0:03:46.17,0:03:49.94,Default,,0000,0000,0000,,And Quividi does a full length thing in \Nenvironments where you have cameras and
Dialogue: 0,0:03:49.94,0:03:54.02,Default,,0000,0000,0000,,you have digital signage. \NSo we have advertisements on screens or
Dialogue: 0,0:03:54.02,0:03:57.70,Default,,0000,0000,0000,,you have information on screens, then \NQuividi clearly allows you to monitor who
Dialogue: 0,0:03:57.70,0:04:01.16,Default,,0000,0000,0000,,is actually watching what you are \Nshowing.
Dialogue: 0,0:04:01.16,0:04:04.69,Default,,0000,0000,0000,,So if you have a bunch of people in front \Nof this camera, it will identify also
Dialogue: 0,0:04:04.69,0:04:08.22,Default,,0000,0000,0000,,people it will say oh, here is a lady, \Nhere's ladies, you also got, a few of the
Dialogue: 0,0:04:08.22,0:04:12.91,Default,,0000,0000,0000,,people are guys. \NIt will make some statistics, how long
Dialogue: 0,0:04:12.91,0:04:16.60,Default,,0000,0000,0000,,the people actually watch for example in \Nadvertisement, where they look on the
Dialogue: 0,0:04:16.60,0:04:21.55,Default,,0000,0000,0000,,screen and so on. \NAnd this entire system is distributed in
Dialogue: 0,0:04:21.55,0:04:26.34,Default,,0000,0000,0000,,the cloud, and Allows you to do a \Ndashboard, a so-called dashboard, of how
Dialogue: 0,0:04:26.34,0:04:31.58,Default,,0000,0000,0000,,your advertisement is being seen in these \Npublic screens, or in the malls where the
Dialogue: 0,0:04:31.58,0:04:39.39,Default,,0000,0000,0000,,screens are being shown. \NAnd at latest, they have 150 networks of
Dialogue: 0,0:04:39.39,0:04:45.33,Default,,0000,0000,0000,,measurements that are deployed all across \Nthe world as you can see.
Dialogue: 0,0:04:45.33,0:04:50.57,Default,,0000,0000,0000,,And there are some very famous names that \Nshow up and so they essentially can do
Dialogue: 0,0:04:50.57,0:04:55.11,Default,,0000,0000,0000,,monitoring of the quality of \Nadvertisement for all of these companies
Dialogue: 0,0:04:55.11,0:04:59.96,Default,,0000,0000,0000,,essentially in real time and provide \Nreports to the effectiveness of using
Dialogue: 0,0:04:59.96,0:05:08.68,Default,,0000,0000,0000,,advertising on screens in public spaces. \NOkay, that's the story of Quividi.
Dialogue: 0,0:05:08.68,0:05:11.58,Default,,0000,0000,0000,,It's cool technology. \NIt uses computer vision, image
Dialogue: 0,0:05:11.58,0:05:15.22,Default,,0000,0000,0000,,processing, the, it also uses a lot of, \Nyou know, state of the art, algorithmic
Dialogue: 0,0:05:15.22,0:05:20.34,Default,,0000,0000,0000,,and machine learning technology. \NPlease visit their website if you want to
Dialogue: 0,0:05:20.34,0:05:26.20,Default,,0000,0000,0000,,know more about this one. \NThe third company is called Sensorscope.
Dialogue: 0,0:05:26.20,0:05:30.81,Default,,0000,0000,0000,,It grew out of all the efforts of doing \Nenvironmental monitoring and various
Dialogue: 0,0:05:30.81,0:05:34.93,Default,,0000,0000,0000,,projects here at DPFL. \NSo, if you want to do environmental
Dialogue: 0,0:05:34.93,0:05:38.35,Default,,0000,0000,0000,,monitoring, it's cool if you can do real \Ntime visualization of what is happening
Dialogue: 0,0:05:38.35,0:05:42.20,Default,,0000,0000,0000,,there. \Nan application where people are very
Dialogue: 0,0:05:42.20,0:05:46.55,Default,,0000,0000,0000,,interesting is so-called precision \Nagriculture, so we want to control the
Dialogue: 0,0:05:46.55,0:05:51.23,Default,,0000,0000,0000,,quality, let's say for example, of water \Nsystems.
Dialogue: 0,0:05:51.23,0:05:55.99,Default,,0000,0000,0000,,you also want to detect you know, certain \Nweather patterns and so on and then
Dialogue: 0,0:05:55.99,0:06:00.73,Default,,0000,0000,0000,,optimize crop production thanks to this \Nmonitoring.
Dialogue: 0,0:06:01.75,0:06:06.18,Default,,0000,0000,0000,,So the company does large scale sensor \Nnetworks, deployments and data
Dialogue: 0,0:06:06.18,0:06:10.37,Default,,0000,0000,0000,,management. \NSo you need wireless sensor networks.
Dialogue: 0,0:06:10.37,0:06:14.47,Default,,0000,0000,0000,,So these are small stations that talk to \Neach other in an ad hoc fashion.
Dialogue: 0,0:06:14.47,0:06:17.76,Default,,0000,0000,0000,,So self organize sensor networks. \NThen you need signal and image
Dialogue: 0,0:06:17.76,0:06:20.31,Default,,0000,0000,0000,,processing. \NThe usual stuff that you have learned
Dialogue: 0,0:06:20.31,0:06:22.83,Default,,0000,0000,0000,,here in the class. \NAnd of course radio communication
Dialogue: 0,0:06:22.83,0:06:27.60,Default,,0000,0000,0000,,technology. \NSo here would be a typical example.
Dialogue: 0,0:06:27.60,0:06:32.28,Default,,0000,0000,0000,,you build a monitoring station. \NWe have seen such monitoring stations at
Dialogue: 0,0:06:32.28,0:06:37.01,Default,,0000,0000,0000,,class when we have discussed sampling \Nissues with respect to rain monitoring.
Dialogue: 0,0:06:37.01,0:06:42.21,Default,,0000,0000,0000,,So you take state-of-the-art, \Noff-the-shelf sophisticated monitoring
Dialogue: 0,0:06:42.21,0:06:46.26,Default,,0000,0000,0000,,communication and so on. \NYou build sensor stations.
Dialogue: 0,0:06:46.26,0:06:49.71,Default,,0000,0000,0000,,You deploy them in a self-organized \Nnetwork.
Dialogue: 0,0:06:49.71,0:06:52.87,Default,,0000,0000,0000,,Then from a bay station you talk to the \Ncloud.
Dialogue: 0,0:06:52.87,0:06:56.81,Default,,0000,0000,0000,,On the cloud, you have all this data, and \Npeople that are interested in monitoring
Dialogue: 0,0:06:56.81,0:07:00.41,Default,,0000,0000,0000,,this sort of deployment get access, \Nprivileged access to this data and can
Dialogue: 0,0:07:00.41,0:07:04.01,Default,,0000,0000,0000,,take statistics and, you know, decide \Nwhat to do, for example, about their
Dialogue: 0,0:07:04.01,0:07:12.06,Default,,0000,0000,0000,,precision agriculture project. \NIt's a small company, half a dozen
Dialogue: 0,0:07:12.06,0:07:16.20,Default,,0000,0000,0000,,people, and probably its main market is \Nprecision all, agriculture, even if it
Dialogue: 0,0:07:16.20,0:07:19.92,Default,,0000,0000,0000,,started, also from a, academic point of \Nview, mostly about environmental
Dialogue: 0,0:07:19.92,0:07:25.30,Default,,0000,0000,0000,,monitoring. \NAnd you can watch their website here, you
Dialogue: 0,0:07:25.30,0:07:29.45,Default,,0000,0000,0000,,can also watch all the data that is \Nonline at climaps.com.
Dialogue: 0,0:07:29.45,0:07:33.30,Default,,0000,0000,0000,,So all the deployments that have ever \Nbeen done by the company and by the lab
Dialogue: 0,0:07:33.30,0:07:37.51,Default,,0000,0000,0000,,are actually available here on, on this \Nwebsite, and you can also use this data
Dialogue: 0,0:07:37.51,0:07:41.60,Default,,0000,0000,0000,,and you know, do some further signal \Nprocessing if you're actually interested
Dialogue: 0,0:07:41.60,0:07:49.65,Default,,0000,0000,0000,,by this topic. \NThe fourth company here is called
Dialogue: 0,0:07:49.65,0:07:53.52,Default,,0000,0000,0000,,Vidinoti. \NIt's a recent start up from the lab and
Dialogue: 0,0:07:53.52,0:07:56.100,Default,,0000,0000,0000,,it works in augmented reality, in \Nparticle augmented reality on mobile
Dialogue: 0,0:07:56.100,0:08:00.69,Default,,0000,0000,0000,,devices. \NSo, the core technologies image
Dialogue: 0,0:08:00.69,0:08:05.72,Default,,0000,0000,0000,,recognition, computer vision. \NBut in a ways that is robust and then to
Dialogue: 0,0:08:05.72,0:08:10.68,Default,,0000,0000,0000,,also do all these processing in real time \Non small devices, like mobile phones and
Dialogue: 0,0:08:10.68,0:08:15.06,Default,,0000,0000,0000,,decide how much processing you do on a \Nmobile phone, how much you do in the
Dialogue: 0,0:08:15.06,0:08:23.19,Default,,0000,0000,0000,,Cloud or on the server. \NAnd Vidinoti has a bunch of state of the
Dialogue: 0,0:08:23.19,0:08:27.27,Default,,0000,0000,0000,,art. \NAlgorithms on the one side, to do
Dialogue: 0,0:08:27.27,0:08:31.72,Default,,0000,0000,0000,,tracking and recognition, and also a \Nnumber of cool ideas on how to do
Dialogue: 0,0:08:31.72,0:08:37.18,Default,,0000,0000,0000,,augmented reality based on these \Nmethodologies.
Dialogue: 0,0:08:40.51,0:08:44.33,Default,,0000,0000,0000,,So, the technology is essentially \Ncloud-based.
Dialogue: 0,0:08:44.33,0:08:48.48,Default,,0000,0000,0000,,But there is an iPhone application that \Nyou can download, and then you can
Dialogue: 0,0:08:48.48,0:08:52.97,Default,,0000,0000,0000,,annotate your favorite pictures or \Nnewspapers or whatever with augmented
Dialogue: 0,0:08:52.97,0:08:57.59,Default,,0000,0000,0000,,reality, and this is actually being used \Nin particular in the newspaper industry
Dialogue: 0,0:08:57.59,0:09:01.81,Default,,0000,0000,0000,,to sort of bring digital content in a, \Nyou know, in a funny or attractive way
Dialogue: 0,0:09:01.81,0:09:08.82,Default,,0000,0000,0000,,onto a medium. \NMainly the newspaper that is being
Dialogue: 0,0:09:08.82,0:09:12.28,Default,,0000,0000,0000,,challenged by, of course by Internet \Ncurrently.
Dialogue: 0,0:09:12.28,0:09:17.92,Default,,0000,0000,0000,,It is a small company, less than 10 \Npeople currently.
Dialogue: 0,0:09:17.92,0:09:22.13,Default,,0000,0000,0000,,It has you know, a strong research and \Ndevelopment.
Dialogue: 0,0:09:22.13,0:09:26.49,Default,,0000,0000,0000,,It has also a strong intellectual \Nportfolio based on patterns that has been
Dialogue: 0,0:09:26.49,0:09:30.69,Default,,0000,0000,0000,,generated over the year around Augmented \NReality.
Dialogue: 0,0:09:32.24,0:09:36.02,Default,,0000,0000,0000,,And if you want to know more, here's a \Nweb site, and here is interactive
Dialogue: 0,0:09:36.02,0:09:40.30,Default,,0000,0000,0000,,application for the iPhone currently, it \Nwill be ported to Android within a couple
Dialogue: 0,0:09:40.30,0:09:47.13,Default,,0000,0000,0000,,of months as well. \NSo, these were example of start-ups that
Dialogue: 0,0:09:47.13,0:09:50.70,Default,,0000,0000,0000,,used state of the art signal processing, \Nimage processing, computer vision, and so
Dialogue: 0,0:09:50.70,0:09:55.84,Default,,0000,0000,0000,,on. \NAnd bring it to the real world, in very
Dialogue: 0,0:09:55.84,0:10:02.18,Default,,0000,0000,0000,,concrete applications, from audio, to \Nsensor networks, to augmented reality, to
Dialogue: 0,0:10:02.18,0:10:07.66,Default,,0000,0000,0000,,monitoring of audience in, advertising.