How To Complete Paint Chip Sample Collection: Fill & Download for Free

I’m glad when people are suspicious and skeptical of our giant corrupt zoo of a government. And I encourage everyone to question the government.Question everything! Why not?However, there are THREE 100% POSITIVE PROOFS that the Apollo missions actually landed on the moon.In order to be 100% positive proof, the evidence needs to COME FROM AN INDEPENDENT SOURCE, such as China, Japan, Or Russia, or it needs to be able to be VERIFIED AS TRUE by an INDEPENDENT SOURCE.The following proof meets that criteria:1. The surface of the moon is currently being 3D mapped by satellites, and the surface mapping data exactly matches the photos the astronauts took.The moon-hoax people say that the landings were faked on a sound stage and the backgrounds were painted by a Hollywood special effects artist. Like this:The problem with that theory is that the artist would not have been able to paint the exact shape of the hills and valleys at the actual lunar sites, because that information was unknown.The technology for gathering three dimensional relief mapping of the lunar surface did not exist in 1969. So the artist’s rendering would have had to come from his own imagination.However, the ability to map the lunar surfaces does exist NOW. And if you feed the new satellite mapping data into a 3D graphic rendering program, the resulting surface EXACTLY MATCHES the Apollo pictures.See the surface contours in this photo, the hills and craters? The 3D data being gathered by LRO satellites exactly matches these contours.In 2008, the Japan Aerospace Exploration Agency (JAXA) SELENE lunar probe obtained several photographs of the Apollo 15 EVA site at station 9A near Hadley Rille.This photo of the lunar surface was taken on the lunar surface by the Apollo 15 astronauts on August 2, 1971 during EVA 3 at station 9A near Hadley Rille:This is a 3D computer graphic reconstruction from 3D data stereo images taken by the Japan Aerospace Exploration Agency (JAXA) SELENE terrain camera and 3D projected to the same vantage point as the surface photos. The background terrain is an exact match with the Apollo photograph (within the SELENE camera resolution of 10 meters).Here are the images side-by-side. They match exactly:This reconstruction can be done for all of the thousands of pictures that the astronauts took from hundreds of different locations on the lunar surface.THIS IS 100% POSITIVE INDEPENDENT EVIDENCE THAT WE LANDED ON THE MOON, AND IT IS IMPOSSIBLE TO FAKE.2. The dirt flying off of the lunar rover wheels flies in a pattern that can only be done on the moon.Suppose you shoot an artillery shell on earth. It will follow a path dictated by air resistance and Earth gravity, which can be precisely calculated. The shape of the path (trajectory) will be sort of lopsided, because of the air resistance slowing the projectile down.But if you shoot the artillery shell on the moon, it will follow a different shaped trajectory due to the lack of air resistance and lower gravity (1/6th of Earth), which can also be precisely calculated. The trajectory shape will be a perfect parabola because of the lack of air resistance. And the projectile will fly way farther that it would on earth due to the lower gravity.This kind of calculation can be applied to the dust flying off of the lunar rover wheels.In films of the lunar landings, the lunar soil flies off of the buggy wheels in a trajectory that can only happen in a vacuum and at 1/6 G gravity.On earth, the dust will be swirled into a vortex dust cloud by the air passing over the rover and will settle back to the earth at much different speed due to the air resistance and the fact that earth gravity is 6 times as strong as the moon’s gravity.Any physicist on earth can take the Apollo films and measure the trajectory of the lunar soil spraying from the wheels and back-calculate to show that it’s in a vacuum, and the gravity force on the dust is the moon gravity of 1/6 G. This calculation has already been done many times.THIS EVIDENCE THAT WE LANDED ON THE MOON IS EASILY VERIFIED BY THE WORLD SCIENTIFIC COMMUNITY AND IS IMPOSSIBLE TO FAKE.This calculation can also be done with several other trajectories of objects on the lunar surface as well:3. NASA has thousands of lunar rocks which have been loaned out to scientists all over the world.These lunar rocks have internal structure that can only be formed on the moon, in an environment with extremely low moisture, low gravity and no atmosphere to screen out radiation called “solar wind.”There is virtually no doubt among the world scientific community that the Apollo lunar samples are genuine moon rock.These rocks have been studied and tested for decades and they are the primary source of scientific data on old the moon is, what its composition is, and how it was formed. It’s amazing how much data the rocks provided.Recently, there has been a lot of excitement in the scientific community over the fact that NASA is going to open three containers of rocks they sealed in 1971.Sealed Cache of Moon Rocks to Be Opened by NASASince they can’t say the rock samples are fake, the moon-hoax people instead say that the moon rock samples are actually lunar meteorites collected on earth in the deserts of Africa and in Antarctica.It’s true that lunar rock has been blasted off the surface of the moon by meteorite impacts and eventually fallen to earth as lunar meteorites.However, these lunar meteorites are INCREDIBLY RARE and HARD TO FIND. The lunar meteorites are scattered far and wide and extremely difficult to spot, as they look identical to earth rock. In all of recorded history only 306 lunar rocks have been found on earth. They are incredibly valuable, even a few small rock chips have sold for hundreds of thousands of dollars.But in the period of 1969 to 1972, NASA gathered a miraculous supply of 2,200 large lunar rocks, which the astronauts brought back from the moon.It took hundreds of years to find 306 lunar rocks on earth, but the conspiracy theorists would have you believe that NASA sent out some government employees with Easter baskets to the African deserts and Antarctica and in a few years they found SEVEN TIMES that amount, including some giant ones.“OK, men, NASA wants us to go to the Antarctica and find a few thousand lunar meteorites worth several billion dollars. Should take a few days at the most. Also, don’t tell anyone. The moon landing hoax is a total secret, only a few thousand aerospace contractors and scientists know about it. And our mortal enemies, the Russians. And now you.”Still not convinced?Well, lunar meteorites on earth are different from the Apollo samples taken by the astronauts. The lunar meteorites have a discoloration on the outside called “fusion crust” that forms when the meteor hits the earth’s atmosphere and gets baked at 3,000 degrees Fahrenheit by the friction. And a lot of the mass of the rock is burned away, including all the sharp edges. And the rock is then weathered by earth rain and wind for thousands of years.So any competent geologist can immediately tell the difference between the Apollo lunar samples and lunar meteorites, they are completely different in appearance and other properties.In addition, many of the Apollo lunar samples are SOIL samples, which cannot enter the earth’s atmosphere without being burned up completely.At any given time, NASA has about 400 moon rocks out on loan to the world scientific community. There is no doubt among the scientific community that these rocks are genuine and were gathered by astronauts.THIS EVIDENCE THAT WE LANDED ON THE MOON IS EASILY VERIFIED BY THE WORLD SCIENTIFIC COMMUNITY AND IS IMPOSSIBLE TO FAKE.In case you’re curious as to how the moon rock samples were collected, here’s a film of an Apollo astronaut Charlie Duke picking up the largest moon rock bought back by the Apollo program (nicknamed “Big Muley.”)Big Muley is a football sized, 26 pound lump of moon breccia rock that was gathered at the edge of Plum Crater in the Descartes Highlands by astronaut Charlie Duke during the Apollo 16 EVA.As I mentioned above, all that is needed to completely refute the entire moon-hoax-conspiracy theory is to show ONE SINGLE irrefutable piece of evidence that we DID go to the moon.Not only is the above evidence 100% verifiable and irrefutable, it has already been verified by third-party scientists from outside the USA.There is also a vast amount of “softer” proof that we went to the moon.By “softer,” I mean stuff that any reasonable person would accept as evidence, but is not 100% irrefutable proof. This is the kind of proof that the police call “corroborating evidence.”For example -US military contractors and aerospace companies actually built the rocket and all the assorted hardware needed to go to the moon. 400,000 people worked 10 years to make a $100 Billion dollar rocket system and equipment to go to the moon.If you keep researching it, you’ll find that the development of the rockets and gear needed to go to the moon was meticulously documented and photographed, and every single little bit of the required hardware was… actually designed and tested and worked.For example, to test the escape abort rocket, they build a small rocket called the Little Joe. And they fired it off and tested it. They made a working escape system.And you can actually go see the escape system in a museum.All the Apollo hardware was all meticulously documented and tested on earth and in space and shown to work. The rockets worked, the spacesuits worked, the onboard computers worked, the communication system worked, the launch complex worked.They had to invent hundreds of new materials and devices and those things spawned thousands of new consumer products.The hardware itself is on public display and the engineers who build it can explain how all of it was developed and how it works.100,000 people stood a few miles away from the launch site in Florida and watched the 5 million pound Saturn V booster take off and disappear into the sky.Why would the USA build an astoundingly advanced rocket and lander and other space hardware capable of going to the moon… and train all the personnel… and then decide not use it?The moon hoax people say NASA didn’t go to the moon because the Van Allen belt radiation would have killed the astronauts.The moon hoax people say that the moon landing is fake because it’s impossible for an astronaut to survive a trip through the Van Allen radiation belts that circle the Earth.But the scientist who DISCOVERED the radiation belts says that the moon hoax people are wrong."The recent Fox (moon hoax conspiracy) TV show, which I saw, is an ingenious and entertaining assemblage of nonsense. The claim that radiation exposure during the Apollo missions would have been fatal to the astronauts is only one example of such nonsense."-- Dr. James Van AllenThe moon hoax guys are literally arguing with the physicist James Van Allen, who the Van Allen belts are named for. How absurd is that?So, if it’s impossible for anyone to travel through the Van Allen Belts, then why wouldn’t the USA just admit it and save all the time and expense of building the Apollo rockets and hardware?Well, the conspiracy theorists say that the USA faked going to the moon for political reasons, to beat the USSR there.It is true that the USA landed on the moon for political reasons: it was during the Cold War and we had to show the world that the USA was stronger than the USSR, by showing that our industrial-military complex was better and stronger.But if the Van Allen Belts were impossible to get through, then NOBODY could have been able to go to the moon.So why would the USA have spent 10+ years and BILLIONS of dollars to develop space hardware to beat the USSR in a race to the moon that NOBODY COULD WIN?It makes NO SENSE AT ALL. It’s simply completely absurd on every level.Of course we can’t forget JFK’s famous speech of May 25, 1961:“I challenge 400,000 of my fellow Americans to spend 10 years working on an astoundingly well engineered $100 billion dollar lunar exploration program. And then fake the landing on a Hollywood sound stage. Why? I don’t know. To fool the public for some reason. The Russians will be utterly humiliated, but they won’t say anything about it being a hoax. Whatever. Who cares? Let’s just build all the equipment and train all the personnel to go to the moon and then not go. For no reason at all.”There was no valid reason to NOT use the rocket to go to the moon.And if there had been a valid reason, then we would have just said so and not built the thing, it cost $100 billion dollars.Another corroborating piece of evidence is the fact that it was possible to track the Apollo spacecraft’s radio transmissions all the way to the moon and back.When it’s pointed out to the moon hoax people that the Russians tracked the Apollo crew’s radio signals all the way to the moon, which counts as third-party verification that the landing was real, the moon-hoax people say “The Russians were in on the conspiracy.”The conspiracy people say that the mighty Saturn V rocket never went to the moon, it just went into low earth orbit for a few days and then the astronauts came back down and said they went to the moon.BUT THE RUSSIANS COULD EASILY TRACK THE LOCATION OF THE SPACECRAFT BY TRACKING THE SOURCE OF THE RADIO COMMUNICATIONS.The Russians could have easily proved that the Apollo rocket never left earth orbit.Forget about the Russians, there were amateur HAM radio guys building directional antenna and tracking the Apollo rocket to the moon and back.If the Apollo spacecraft had never left Earth orbit, the Russians could have easily proved it. So why on earth would the Russians help us to utterly humiliate them before the whole world, during the height of the Cold War?The moon hoax people say it’s because the Russians “needed US wheat to survive.”The idea is that the Russians had a bad wheat harvest and were on the brink of starvation, so the USA blackmailed then with US wheat into letting the US win the fake moon race.We blackmailed the USSR with wheat!“Natasha, eat the Frosted Mini Wheats given to us by the American imperialist dogs and their capitalist lackeys. We lost the war for global domination but we gained a tasty way to start our day off right with a nutritious breakfast.”This theory is complete nonsense and it can be easily disproved by looking at agricultural data from 1970. If anyone cares to bother.Remember that in 1969 the USA and USSR were locked in a death battle for global domination.This was not a joke. The USA fought the Korean War and Vietnam War to stop the spread of Russian and Chinese backed Communism. This was serious business and a VAST NUMBER of people died.The world was literally poised on the brink of nuclear extermination of the entire human race.Every adult US male in the 1950s through the late 1980s sort of thought he might eventually die on a radioactive battlefield in Europe, fighting the Soviet Bloc.So the moon hoax conspiracy idea is that the USSR was spending hundreds of billions of dollars on an enormous modern army with tanks and troops and hundreds of navy ships and nuclear submarines and missile silos, but they couldn’t afford to buy WHEAT from China or France or Argentina.The Russian sailors on board this huge and incredibly expensive nuclear submarine that could vaporize the entire country of Germany with nuclear missiles were VERY HUNGRY. They had no Wheat in the pantry.So the USSR had to admit defeat in a fake moon race and be completely and totally humiliated by the USA before the rest of the world, and maybe end up being exterminated by the USA by losing the political support of the fence-sitters on the sidelines who were waiting to see if the USA or USSR was stronger.Because of wheat.Are you by any chance starting to see how goofy that the moon-hoax conspiracy theory is? It simply makes no sense on any level.When the Apollo astronauts visited the moon, in addition to bringing back rock samples, they left behind part of their spacecraft and many science experiment devices.The above picture shows the bottom part of the lunar lander called the “descent stage” which is still on the moon. This photo was taken by the camera mounted on the moon buggy. The moon buggy camera was used to take a video of the blastoff of the ascent stage of the lander when the astronauts left the moon:That descent stage is still on the moon, and so is the moon buggy.Any nation who wants to can fly a satellite over the Apollo landing sites and see the equipment there.In addition, the astronauts left behind scientific experiments which sent back data that is currently being used by scientists all over the earth.The only experimental equipment still sending back information to the earth are the laser beam reflectors. These reflectors are used to bounce lasers beams back to the earth, and the data is used to precisely calculate the moon’s position.There is no doubt whatsoever among the world’s scientists that the astronauts landed on the moon.In contrast to the above evidence that astronauts went to the moon, some of which is irrefutable, consider how flimsy the moon-hoax “evidence” always is.The moon hoax conspiracy “evidence” is never scientific, it’s ALWAYS eventually refuted by the international science community.The original moon hoax denier was a technical writer for a defense contractor. He lived in a trailer with 37 cats and thought that the CIA was out to kill him.The stuff he said about the Apollo programs is plain old goofy. Check it out:That’s the kind of credentials that the moon hoax people typically have. They’re not scientists, they’re conspiracy crackpots.“My cat Fluffy says that Neil Armstrong was a clone and the original Neil Armstrong is frozen in cryo-stasis. Along with Hitler.”Every time these moon hoax guys dreams up with a new piece of “evidence”, a scientist comes along and debunks it.The original “shocking” evidence that the moon landings were a hoax was the idea that there are no stars in the background of the lunar landing photos. This “evidence” was screeched and screeched about by the moon landing conspiracy people for a few years.Where are the stars in this photo? It must be fake! They forgot to paint stars in the background!But any high school photography buff can tell you that it’s IMPOSSIBLE TO MAKE A CAMERA TAKE A PICTURE OF A BRILLIANTLY SUNLIT LUNAR SURFACE AND AND ALSO A STAR IN THE BACKGROUND.The camera closes the light aperture to adjust to the sunlight and therefore the stars won’t show up in the picture. You can actually prove this tonight: try to take a picture of a lit-up street and the stars in the sky at the same time.Hey! Where are the stars in this photo? This street must be a hoax!This fact was finally pounded into most of (but not all) the moon hoaxers brains, so they got more sophisticated about it. Now the moon hoax people say that the shadow side of the lunar lander was lit up too bright in the photos, it should be darker shadow.So some computer experts did an incredibly sophisticated computer simulation to show that the moon hoax guys were dead wrong:Meticulous Visual Recreation Of Moon Landing Shows It Wasn't A HoaxOh well, another bit of moon hoax “evidence” gets deflated by actual experts. Cue sad trombone music.One by one, all the moon hoax theories are shot down in this way.Most of the moon hoax conspiracy “evidence” is just empty speculation and accusation.The moon hoax guys pointed out that the Apollo astronauts looked glum and exhausted after they returned to earth, and stated that this was evidence that they were lying about going to the moon.As if the astronauts couldn’t possibly be suffering from nervous and physical exhaustion after a 10 day semi-suicide mission in a spaceship the size of a compact car. And then immediately spending a few weeks in quarantine.This is the kind of empty speculation and empty accusation that the moon hoax guys think of as valid evidence.Scientifically, the moon hoax conspiracy ranks right up there with Bigfoot and the Loch Ness Monster.Actually, that’s not true. The people who believe in Bigfoot have more intellectual standing than the moon hoax people.To be on par with moon hoax guys, the Bigfoot people would have to claim that the US government used 400,000 defense contractor employees and spent $100 billion to build Bigfoot a rocket with 7.5 million pounds of thrust as part of a hoax, and thousands of scientists and engineers around the world knew about it and covered it up, including the USSR.THEN the Bigfoot people would be as absurd as the moon hoax people.So why won’t the moon hoax people shut up and go look for Bigfoot instead?Because it’s very fun and emotionally satisfying to have a conspiracy theory.The love of conspiracy theories is hard-wired into our brain structure. It’s fun and interesting and it makes people feel superior.Why Do Some People Believe in Conspiracy Theories?Ever notice the smug attitude of conspiracy theorists? Being a conspiracy theorist is a huge source of self-worth to them. They see themselves as amazingly clever and superior to the average citizen.“Oh, you simple minded fools have been duped into believing an obvious lie. If only you had my amazing powers of intellect you would be able to see right through it.”SUMMARY -I strongly urge you to examine all the moon landing conspiracy theory “evidence” and the debunking of the “evidence” by the international scientific community.See for yourself how completely absurd and bogus that the moon landing hoax conspiracy theory is.But keep in the back of your mind the three irrefutable proofs that we did land on the moon, which completely invalidate the entire conspiracy theory.

Yes, They definitely are Here are few ideas for decorating your dream home like pro by these easy steps,Are all the walls the same color and the ceiling too?What color(s) is the room? Write this down. Is there a chair rail or decorative molding at the ceiling, wall covering or faux painting?What is on the floor? Is there tile, wood, carpet, area rugs? What color and texture and pattern? Is the area rug braided or a Persian rug or a contemporary design diagonally placed partly under one of the chairs and sofa?What style is the furniture? Is it all a matching set?Do all the woods match? Maybe the sofa and club chairs are the same style and wood finish but the chairs have a patterned upholstery and the sofa is a solid or striped fabric. Write this down.How is the furniture arranged? Does the sofa sit in the middle of the room separating the conversation area from the game table, or is all the furniture pushed up against the walls? Draw a little floor plan so you don’t forget when you get home.The windows. Are there blinds, shutters, a valence or draperies? Are the rods large or small, wood or metal with huge finials? Make a note of it. If there are panels at each side of the window, do they go all the way to the floor? What color are they? Do they match the walls or are they the same fabric as the toss pillows on the sofa? Is it the drapery hold back that catches your eye? You will have to consider the exposure and privacy out your own windows when you make your final window treatment selection.The pizzazz is the accessories. This is often what many people love most about the model homes or magazine pictures that inspire them. The details, the little things we all love to collect and buy when we shop. This is what pulls the room together and gives it your personality. In the room you love, what kind of lamps are there? Is there a place to store the things you will need to put away? Are there several things grouped on the coffee table? What sizes are they? Is there something pretty tall, a large round platter set on a stand and a box all grouped together? If there are photographs on the end tables, how big are they and how does the frame style work with everything else in the room?What is on the walls? Is there one large picture over the sofa or is there an arrangement of black and white photos in different size frames? What color is the matting and how much space from the photo to the frame? Details, details, details. Draw a sketch of the picture arrangement and the sizes in your notebook.Congratulations !!. You have just completed the first step to decorating your own home.You know what you want, how it goes together and what it will look like. You don’t have to break the bank now to create your room. You probably already have the furniture pieces and accessories. You may already have the lamps and area rug as well. You may just have to recreate and rearrange.Get started. Go to the paint store and select the paint chips you think best match your dream room. Grab several samples, color can be deceiving and they can look different depending on the light. Take them back to the model home and see which ones match the best. Buy the paint and paint the room.Your window treatment. If you copy your dream room exactly make sure you have the right privacy and exposure considerations covered. This is where a lot of people go astray. They think they like shutters and buy them. Then they don’t like them and also wonder how their dream room went wrong. Your dream room may have been more airy with a shade and drapery panels on large wood poles. The wrong blinds, shutters or draperies and hardware can be a costly mistake that you probably won’t change.Arrange your furniture for conversation, TV watching, family fun, reading, eating or homework keeping in mind your dream room. Was all the furniture lined up on one side of the room? Your sofa may be the wrong color and your chairs may need to be redone. That doesn’t mean you have to buy all new; consider slipcovers. You can buy them or make them.Be sure your accessories are things you want around you and are of a grand enough scale for your space. Voila! You’re on your way to decorating confidence and living in the home that makes you and your family happy.

Deep Learning refers to a class of machine learning (ML) techniques that combine the following:Large neural networks (millions of free parameters)High performance computing ( thousands of processors running in parallel)Big Data (e.g. millions of color images or recorded chess games)Deep learning techniques currently achieve state of the art performance in a multitude of problem domains (vision, audio, robotics, natural language processing, to name a few). Recent advances in Deep Learning also incorporate ideas from statistical learning [1,2], reinforcement learning (RL) [3], and numerical optimization . For a broad survey of the field, see [9,10].In no particular order, here are some product categories made possible with today's deep learning techniques: customized data compression, compressive sensing, data-driven sensor calibration, offline AI, human-computer interaction, gaming, artistic assistants, unstructured data mining, voice synthesis.Customized data compressionSuppose you are designing a video conferencing app and want to come up with a lossy encoding scheme to reduce the number of packets you need to send over the Internet. You could use an off-the-shelf codec like H.264, but H.264 is not optimal because it is calibrated for generic video - anything from cat videos to feature films to clouds. It would be nice if instead we had a video codec that was optimized for specifically FaceTime videos. We can save even more bytes than a generic algorithm if we take advantage of the fact that most of the time, there is a face in the center of the screen. However, designing such an encoding scheme is tricky. How do we specify where the face is positioned, how much eyebrow hair the subject has, what color their eyes are, the shape of their jaw, etc? What if their hair is covering one of their eyes? What if there are zero or multiple faces in the picture?Deep learning can be applied here. Auto-encoders are a type of neural network whose output is merely a copy of the input data. Learning this "identity mapping" would be trivial if it weren't for the fact that the hidden layers of the auto-encoder are chosen to be smaller than the input layer. This "information bottleneck" forces the auto-encoder to learn an compressed representation of the data in the hidden layer, which is then decoded back to the original form by the remaining layers in the network.Through end-to-end training, auto-encoders and other deep learning techniques *adapt* to the specific nuances of your data. Unlike principal components analysis, the encoding and decoding steps are not limited to affine (linear) transformations. PCA learns an "encoding linear transform", while auto-encoders learn a "encoding program".This makes neural nets far more powerful, and allows for complex, domain-specific compression; anything from storing a gazillion selfies on Facebook, to faster YouTube video streaming, to scientific data compression, to reducing the space needed for your personal iTunes library. Imagine if your iTunes library learned a "country music" auto-encoder just to compress your personal music collection!Compressive sensingCompressive sensing is closely related to the decoding aspects of lossy compression. Many interesting signals have a particular structure to them - that is, the distribution of signals is not completely arbitrary. This means that we don't actually have to sample at the Nyquist limit in order to obtain a perfect reconstruction of the signal, as long our decoding algorithm can properly exploit the underlying structure.Deep learning is applicable here because we can use neural networks to learn the sparse structure without manual feature engineering. Some product applications:Super-resolution algorithms (waifu2X)- literally an "enhance" button like those from CSI Miamiusing WiFi radio wave interference to see people through walls (MIT Wi-Vi)interpreting 3D structure of an object given incomplete observations (such as a 2D image or partial occlusionmore accurate reconstructions from sonar / LIDAR dataData-driven sensor calibrationGood sensors and measurement devices often rely on expensive, precision-manufactured components.Take digital cameras, for example. Digital cameras assume the glass lens is of a certain "nice" geometry. When taking a picture, the onboard processor solves the light transport equations through the lens to compute the final image.If the lens is scratched, or warped or shaped like a bunny (instead of a disc) these assumptions are broken and the images no longer turn out well. Another example: our current decoding models used in MRI and EEG assume the cranium is a perfect sphere in order to keep the math manageable [4]. This sort of works, but sometimes we miss the location of a tumor by a few mm. More accurate photographic and MRI imaging ought to compensate for geometric deviation, whether they result from underlying sources or manufacturing defects.Fortunately, deep learning allows us to calibrate our decoding algorithms with data.Instead of a one-size-fits-all decoding model (such as a Kalman filter), we can express more complex biases specifically tuned to each patient or each measuring device. If our camera lens is scratched, we can train the decoding software to implicitly compensate for the altered geometry. This means we no longer have to manufacture and align sensors with utmost precision, and this saves a lot of money.In some cases, we can do away with hardware completely and let the decoding algorithm compensate for that; the Columbia Computational Photography lab has developed a kind of camera that doesn't have a lens. Software-defined imaging, so to speak.Offline AIBeing able to run AI algorithms without Internet is crucial for apps that have low latency requirements (I.e. self driving cars & robotics) or do not have reliable connectivity (smartphone apps for traveling).Deep Learning is especially suitable for this. After the training phase, neural networks can run the feed forward step very quickly. Furthermore, it is straightforward to shrink down large neural nets into small ones, until they are portable enough to run on a smartphone (at the expense of some accuracy).Google has already done this in their offline camera translation feature in Google Translate App [6].Some other possibilities:Intelligent assistants (e.g. Siri) that retain some functionality even when offline.wilderness survival app that tells you if that plant is poison ivy, or whether those mushrooms are safe to eatsmall drones with on-board TPU chips [11] that can perform simple obstacle avoidance and navigationHuman-computer interactionDeep Neural Networks are the first kind of models that can really see and hear our world with an acceptable level of robustness. This opens up a lot of possibilities for Human-Computer Interaction.Cameras can now be used to read sign language and read books aloud to people. In fact, deep neural networks can now describe to us in full sentences what they see [12]. Baidu's DuLight project is enabling visually-impaired people to see the world around them through a sight-to-speech earpiece.Dulight--Eyes for visually impairedWe are not limited to vision-based HCI. Deep learning can help calibrate EEG interfaces for paraplegics to interact with computers more rapidly, or provide more accurate decoding tech for projects like Soli [7].GamingGames are computationally challenging because they run physics simulation, AI logic, rendering, and multiplayer interaction together in real time. Many of these components have at least O(N^2) in complexity, so our current algorithms have hit their Moore's ceiling.Deep learning pushes the boundaries on what games are capable of in several ways.Obviously, there's the "game AI" aspect. In current video games, AI logic for non-playable characters (NPC) are not much more than a bunch of if-then-else statements tweaked to imitate intelligent behavior. This is not clever enough for advanced gamers, and leads to somewhat unchallenging character interaction in single-player mode. Even in multiplayer, a human player is usually the smartest element in the game loop.This changes with Deep Learning. Google Deepmind's AlphaGo has shown us that Deep Neural Networks, combined with policy gradient learning, are powerful enough to beat the strongest of human players at complex games like Go. The Deep Learning techniques that drive AlphaGo may soon enable NPCs that can exploit the player's weaknesses and provide a more engaging gaming experience. Game data from other players can be sent to the cloud for training the AI to learn from its own mistakes.Another application of deep learning in games is physics simulation. Instead of simulating fluids and particles from first principles, perhaps we can turn the nonlinear dynamics problem into a regression problem. For instance, if we train a neural net to learn the physical rules that govern fluid dynamics, we can evaluate it quickly during gameplay without having to perform large-scale solutions to Navier stokes equations in real time.In fact, this has been done already by Ladicky and Jeong 2015 [8].For VR applications that must run at 90 FPS minimum, this may be the only viable approach given current hardware constraints.Third, deep generative modeling techniques can be used to create unlimited, rich procedural content - fauna, character dialogue, animation, music, perhaps the narrative of the game itself. This is an area that is just starting to be explored by games like No Man's Sky, which could potentially make games with endless novel content.To add a cherry on top, Deep Neural nets are well suited for parallel mini-batched evaluation, which means that AI logic for a 128 NPCs or 32 water simulations might be evaluated simultaneously on a single graphics card.Artistic AssistantsGiven how well neural networks perceive images, audio, and text, it's no surprise that they also work when we use them to draw paintings [13], compose music [14], and write fiction [15].People have been trying to get computers to compose music and paint pictures for ages, but deep learning is the first one that actually generates "good results". There are already several apps in the App Store that implement these algorithms for giggles, but soon we may see them as assistive generators/filters in professional content creation software.Data Mining from Unstructured DataDeep learning isn't at the level where it can extract the same amount of information humans can from web pages, but the vision capabilities of deep neural nets are good enough for allowing machines to understand more than just hypertext.For instance:Parsing events from scanned flyersidentifying which products on EBay are the samedetermining consumer sentiment from webcamextracting blog content from pages without RSS feedsintegrate photo information into valuing financial instruments, insurance policies, and credit scores.Voice synthesisGenerative modeling techniques have come far enough and there is sufficient data out there that it is only a matter of time before someone makes an app that reads aloud to you in Morgan Freeman's or Scarlet Johansen's voice. At Vanguard, my voice is my password.Bonus: more productsAdaptive OS / Network stack scheduling - scheduling threads and processes in an OS is a NP hard problem. We don't have a very satisfactory solution to this right now, and scheduling algorithms in modern operating systems, filesystems, and TCP/IP implementations are all fairly simple. Perhaps if a small neural net could be used to adapt to a user's particular scheduling patterns (frame this as an RL problem), we would decrease scheduling overhead incurred by the OS. This might make a lot of sense inside of data centers where the savings can really scale.colony counting & cell tracking for microscopy software (for wet lab research)The strategy of "replacing simulation with machine learning" has been useful in the fields of drug design too, presenting enormous speed ups in finding which compounds are helpful or toxic [untethiner 2015].References[1] Auto encoding variation all bayes [1312.6114] Auto-Encoding Variational Bayes[2] One shot deep generative models https://arxiv.org/pdf/1603.05106v2.pdf[3] Deep Reinforcement Learning: Pong from Pixels[4] http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3790855/[5] http://web.media.mit.edu/~gordonw/courses/ComputationalPlenopticImaging/CPICourseNotes.pdf[6] How Google Translate squeezes deep learning onto a phone[7] Project Soli[8] Data-driven Fluid Simulations using Regression Forests[9] http://www.nature.com/nature/journal/v521/n7553/full/nature14539.html[10] Deep Learning in Neural Networks: An Overview[11] Google supercharges machine learning tasks with TPU custom chip[12] http://arxiv.org/pdf/1410.1090[13] [1508.06576] A Neural Algorithm of Artistic Style[14] Composing Music With Recurrent Neural Networks[15] Writing with the machine