Tag Archives: learn

#437564 How We Won the DARPA SubT Challenge: ...

This is a guest post. The views expressed here are those of the authors and do not necessarily represent positions of IEEE or its organizational units.​

“Do you smell smoke?” It was three days before the qualification deadline for the Virtual Tunnel Circuit of the DARPA Subterranean Challenge Virtual Track, and our team was barrelling through last-minute updates to our robot controllers in a small conference room at the Michigan Tech Research Institute (MTRI) offices in Ann Arbor, Mich. That’s when we noticed the smell. We’d assumed that one of the benefits of entering a virtual disaster competition was that we wouldn’t be exposed to any actual disasters, but equipment in the basement of the building MTRI shares had started to smoke. We evacuated. The fire department showed up. And as soon as we could, the team went back into the building, hunkered down, and tried to make up for the unexpected loss of several critical hours.

Team BARCS joins the SubT Virtual Track
The smoke incident happened more than a year after we first learned of the DARPA Subterranean Challenge. DARPA announced SubT early in 2018, and at that time, we were interested in building internal collaborations on multi-agent autonomy problems, and SubT seemed like the perfect opportunity. Though a few of us had backgrounds in robotics, the majority of our team was new to the field. We knew that submitting a proposal as a largely non-traditional robotics team from an organization not known for research in robotics was a risk. However, the Virtual Track gave us the opportunity to focus on autonomy and multi-agent teaming strategies, areas requiring skill in asynchronous computing and sensor data processing that are strengths of our Institute. The prevalence of open source code, small inexpensive platforms, and customizable sensors has provided the opportunity for experts in fields other than robotics to apply novel approaches to robotics problems. This is precisely what makes the Virtual Track of SubT appealing to us, and since starting SubT, autonomy has developed into a significant research thrust for our Institute. Plus, robots are fun!

After many hours of research, discussion, and collaboration, we submitted our proposal early in 2018. And several months later, we found out that we had won a contract and became a funded team (Team BARCS) in the SubT Virtual Track. Now we needed to actually make our strategy work for the first SubT Tunnel Circuit competition, taking place in August of 2019.

Building a team of virtual robots
A natural approach to robotics competitions like SubT is to start with the question of “what can X-type robot do” and then build a team and strategy around individual capabilities. A particular challenge for the SubT Virtual Track is that we can’t design our own systems; instead, we have to choose from a predefined set of simulated robots and sensors that DARPA provides, based on the real robots used by Systems Track teams. Our approach is to look at what a team of robots can do together, determining experimentally what the best team configuration is for each environment. By the final competition, ideally we will be demonstrating the value of combining platforms across multiple Systems Track teams into a single Virtual Track team. Each of the robot configurations in the competition has an associated cost, and team size is constrained by a total cost. This provides another impetus for limiting dependence on complex sensor packages, though our ranging preference is 3D lidar, which is the most expensive sensor!

Image: Michigan Tech Research Institute

The teams can rely on realistic physics and sensors but they start off with no maps of any kind, so the focus is on developing autonomous exploratory behavior, navigation methods, and object recognition for their simulated robots.

One of the frequent questions we receive about the Virtual Track is if it’s like a video game. While it may look similar on the surface, everything under the hood in a video game is designed to service the game narrative and play experience, not require novel research in AI and autonomy. The purpose of simulations, on the other hand, is to include full physics and sensor models (including noise and errors) to provide a testbed for prototyping and developing solutions to those real-world challenges. We are starting with realistic physics and sensors but no maps of any kind, so the focus is on developing autonomous exploratory behavior, navigation methods, and object recognition for our simulated robots.

Though the simulation is more like real life than a video game, it is not real life. Due to occasional software bugs, there are still non-physical events, like the robots falling through an invisible hole in the world or driving through a rock instead of over it or flipping head over heels when driving over a tiny lip between world tiles. These glitches, while sometimes frustrating, still allow the SubT Virtual platform to be realistic enough to support rapid prototyping of controller modules that will transition straightforwardly onto hardware, closing the loop between simulation and real-world robots.

Full autonomy for DARPA-hard scenarios
The Virtual Track requirement that the robotic agents be fully autonomous, rather than have a human supervisor, is a significant distinction between the Systems and Virtual Tracks of SubT. Our solutions must be hardened against software faults caused by things like missing and bad data since our robots can’t turn to us for help. In order for a team of robots to complete this objective reliably with no human-in-the-loop, all of the internal systems, from perception to navigation to control to actuation to communications, must be able to autonomously identify and manage faults and failures anywhere in the control chain.

The communications limitations in subterranean environments (both real and virtual) mean that we need to keep the amount of information shared between robots low, while making the usability of that information for joint decision-making high. This goal has guided much of our design for autonomous navigation and joint search strategy for our team. For example, instead of sharing the full SLAM map of the environment, our agents only share a simplified graphical representation of the space, along with data about frontiers it has not yet explored, and are able to merge its information with the graphs generated by other agents. The merged graph can then be used for planning and navigation without having full knowledge of the detailed 3D map.

The Virtual Track requires that the robotic agents be fully autonomous. With no human-in-the-loop, all of the internal systems, from perception to navigation to control to actuation to communications, must be able to identify and manage faults and failures anywhere in the control chain.

Since the objective of the SubT program is to advance the state-of-the-art in rapid autonomous exploration and mapping of subterranean environments by robots, our first software design choices focused on the mapping task. The SubT virtual environments are sufficiently rich as to provide interesting problems in building so-called costmaps that accurately separate obstructions that are traversable (like ramps) from legitimately impassible obstructions. An extra complication we discovered in the first course, which took place in mining tunnels, was that the angle of the lowest beam of the lidar was parallel to the down ramps in the tunnel environment, so they could not “see” the ground (or sometimes even obstructions on the ramp) until they got close enough to the lip of the ramp to receive lidar reflections off the bottom of the ramp. In this case, we had to not only change the costmap to convince the robot that there was safe ground to reach over the lip of the ramp, but also had to change the path planner to get the robot to proceed with caution onto the top of the ramp in case there were previously unseen obstructions on the ramp.

In addition to navigation in the costmaps, the robot must be able to generate its own goals to navigate to. This is what produces exploratory behavior when there is no map to start with. SLAM is used to generate a detailed map of the environment explored by a single robot—the space it has probed with its sensors. From the sensor data, we are able to extract information about the interior space of the environment while looking for holes in the data, to determine things like whether the current tunnel continues or ends, or how many tunnels meet at an intersection. Once we have some understanding of the interior space, we can place navigation goals in that space. These goals naturally update as the robot traverses the tunnel, allowing the entire space to be explored.

Sending our robots into the virtual unknown
The solutions for the Virtual Track competitions are tested by DARPA in multiple sequestered runs across many environments for each Circuit in the month prior to the Systems Track competition. We must wait until the joint award ceremony at the conclusion of the Systems Track to find out the results, and we are completely in the dark about placings before the awards are announced. It’s nerve-wracking! The challenges of the worlds used in the Circuit events are also hand-designed, so features of the worlds we use for development could be combined in ways we have not anticipated—it’s always interesting to see what features were prioritized after the event. We test everything in our controllers well enough to feel confident that we at least are submitting something reasonably stable and broadly capable, and once the solution is in, we can’t really do anything other than “let go” and get back to work on the next phase of development. Maybe it’s somewhat like sending your kid to college: “we did our best to prepare you for this world, little bots. Go do good.”

Image: Michigan Tech Research Institute

The first SubT competition was the Tunnel Circuit, featuring a labyrinthine environment that simulated human-engineered tunnels, including hazards such as vertical shafts and rubble.

The first competition was the Tunnel Circuit, in October 2019. This environment models human-engineered tunnels. Two substantial challenges in this environment were vertical shafts and rubble. Our team accrued 21 points over 15 competition runs in five separate tunnel environments for a second place finish, behind Team Coordinated Robotics.

The next phase of the SubT virtual competition was the Urban Circuit. Much of the difference between our Tunnel and Urban Circuit results came down to thorough testing to identify failure modes and implementations of checks and data filtering for fault tolerance. For example, in the SLAM nodes run by a single robot, the coordinates of the most recent sensor data are changed multiple times during processing and integration into the current global 3D map of the “visited” environment stored by that robot. If there is lag in IMU or clock data, the observation may be temporarily registered at a default location that is very far from the actual position. Since most of our decision processes for exploration are downstream from SLAM, this can cause faulty or impossible goals to be generated, and the robots then spend inordinate amounts of time trying to drive through walls. We updated our method to add a check to see if the new map position has jumped a far distance from the prior map position, and if so, we threw that data out.

Image: Michigan Tech Research Institute

In open spaces like the rooms in the Urban circuit, we adjusted our approach to exploration through graph generation to allow the robots to accurately identify viable routes while helping to prevent forays off platform edges.

Our approach to exploration through graph generation based on identification of interior spaces allowed us to thoroughly explore the centers of rooms, although we did have to make some changes from the Tunnel circuit to achieve that. In the Tunnel circuit, we used a simplified graph of the environment based on landmarks like intersections. The advantage of this approach is that it is straightforward for two robots to compare how the graphs of the space they explored individually overlap. In open spaces like the rooms in the Urban circuit, we chose to instead use a more complex, less directly comparable graph structure based on the individual robot’s trajectory. This allowed the robots to accurately identify viable routes between features like subway station platforms and subway tracks, as well as to build up the navigation space for room interiors, while helping to prevent forays off the platform edges. Frontier information is also integrated into the graph, providing a uniform data structure for both goal selection and route planning.

The results are in!
The award ceremony for the Urban Circuit was held concurrently with the Systems Track competition awards this past February in Washington State. We sent a team representative to participate in the Technical Interchange Meeting and present the approach for our team, and the rest of us followed along from our office space on the DARPAtv live stream. While we were confident in our solution, we had also been tracking the online leaderboard and knew our competitors were going to be submitting strong solutions. Since the competition environments are hand-designed, there are always novel challenges that could be presented in these environments as well. We knew we would put up a good fight, but it was very exciting to see BARCS appear in first place!

Any time we implement a new module in our control system, there is a lot of parameter tuning that has to happen to produce reliably good autonomous behavior. In the Urban Circuit, we did not sufficiently test some parameter values in our exploration modules. The effect of this was that the robots only chose to go down small hallways after they explored everything else in their environment, which meant very often they ran out of time and missed a lot of small rooms. This may be the biggest source of lost points for us in the Urban Circuit. One of our major plans going forward from the Urban Circuit is to integrate more sophisticated node selection methods, which can help our robots more intelligently prioritize which frontier nodes to visit. By going through all three Circuit challenges, we will learn how to appropriately add weights to the frontiers based on features of the individual environments. For the Final Challenge, when all three Circuit environments will be combined into large systems, we plan to implement adaptive controllers that will identify their environments and use the appropriate optimized parameters for that environment. In this way, we expect our agents to be able to (for example) prioritize hallways and other small spaces in Urban environments, and perhaps prioritize large openings over small in the Cave environments, if the small openings end up being treacherous overall.

Next for our team: Cave Circuit
Coming up next for Team BARCS is the Virtual Cave Circuit. We are in the middle of testing our hypothesis that our controller will transition from UGVs to UAVs and developing strategies for refining our solution to handle Cave Circuit environmental hazards. The UAVs have a shorter battery life than the UGVs, so executing a joint exploration strategy will also be a high priority for this event, as will completing our work on graph sharing and merging, which will give our robot teams more sophisticated options for navigation and teamwork. We’re reaching a threshold in development where we can start increasing the “smarts” of the robots, which we anticipate will be critical for the final competition, where all of the challenges of SubT will be combined to push the limits of innovation. The Cave Circuit will also have new environmental challenges to tackle: dynamic features such as rock falls have been added, which will block previously accessible passages in the cave environment. We think our controllers are well-poised to handle this new challenge, and we’re eager to find out if that’s the case.

As of now, the biggest worries for us are time and team composition. The Cave Circuit deadline has been postponed to October 15 due to COVID-19 delays, with the award ceremony in mid-November, but there have also been several very compelling additions to the testbed that we would like to experiment with before submission, including droppable networking ‘breadcrumbs’ and new simulated platforms. There are design trade-offs when balancing general versus specialist approaches to the controllers for these robots—since we are adding UAVs to our team for the first time, there are new decisions that will have to be made. For example, the UAVs can ascend into vertical spaces, but only have a battery life of 20 minutes. The UGVs by contrast have 90 minute battery life. One of our strategies is to do an early return to base with one or more agents to buy down risk on making any artifact reports at all for the run, hedging against our other robots not making it back in time, a lesson learned from the Tunnel Circuit. Should a UAV take on this role, or is it better to have them explore deeper into the environment and instead report their artifacts to a UGV or network node, which comes with its own risks? Testing and experimentation to determine the best options takes time, which is always a worry when preparing for a competition! We also anticipate new competitors and stiffer competition all around.

Image: Michigan Tech Research Institute

Team BARCS has now a year to prepare for the final DARPA SubT Challenge event, expected to take place in late 2021.

Going forward from the Cave Circuit, we will have a year to prepare for the final DARPA SubT Challenge event, expected to take place in late 2021. What we are most excited about is increasing the level of intelligence of the agents in their teamwork and joint exploration of the environment. Since we will have (hopefully) built up robust approaches to handling each of the specific types of environments in the Tunnel, Urban, and Cave circuits, we will be aiming to push the limits on collaboration and efficiency among the agents in our team. We view this as a central research contribution of the Virtual Track to the Subterranean Challenge because intelligent, adaptive, multi-robot collaboration is an upcoming stage of development for integration of robots into our lives.

The Subterranean Challenge Virtual Track gives us a bridge for transitioning our more abstract research ideas and algorithms relevant to this degree of autonomy and collaboration onto physical systems, and exploring the tangible outcomes of implementing our work in the real world. And the next time there’s an incident in the basement of our building, the robots (and humans) of Team BARCS will be ready to respond.

Richard Chase, Ph.D., P.E., is a research scientist at Michigan Tech Research Institute (MTRI) and has 20 years of experience developing robotics and cyber physical systems in areas from remote sensing to autonomous vehicles. At MTRI, he works on a variety of topics such as swarm autonomy, human-swarm teaming, and autonomous vehicles. His research interests are the intersection of design, robotics, and embedded systems.

Sarah Kitchen is a Ph.D. mathematician working as a research scientist and an AI/Robotics focus area leader at MTRI. Her research interests include intelligent autonomous agents and multi-agent collaborative teams, as well as applications of autonomous robots to sensing systems.

This material is based upon work supported by the Defense Advanced Research Projects Agency (DARPA) under Contract No. HR001118C0124 and is released under Distribution Statement (Approved for Public Release, Distribution Unlimited). Any opinions, findings and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of DARPA. Continue reading

Posted in Human Robots

#437504 A New and Improved Burger Robot’s on ...

No doubt about it, the pandemic has changed the way we eat. Never before have so many people who hated cooking been forced to learn how to prepare a basic meal for themselves. With sit-down restaurants limiting their capacity or shutting down altogether, consumption of fast food and fast-casual food has skyrocketed. Don’t feel like slaving over a hot stove? Just hit the drive through and grab a sandwich and some fries (the health implications of increased fast food consumption are another matter…).

Given our sudden immense need for paper-wrapped burgers and cardboard cartons of fries, fast food workers are now counted as essential. But what about their safety, both from a virus standpoint and from the usual risks of working in a busy kitchen (like getting burned by the stove or the hot oil from the fryer, cut by a slicer, etc.)? And how many orders of burgers and fries can humans possibly churn out in an hour?

Enter the robot. Three and a half years ago, a burger-flipping robot aptly named Flippy, made by Miso Robotics, made its debut at a fast food restaurant in California called CaliBurger. Now Flippy is on the market for anyone who wishes to purchase their own, with a price tag of $30,000 and a range of new capabilities—this burger bot has progressed far beyond just flipping burgers.

Flippy’s first iteration was already pretty impressive. It used machine learning software to locate and identify objects in front of it (rather than needing to have objects lined up in specific spots), and was able to learn from experience to improve its accuracy. Sensors on its grill-facing side took in thermal and 3D data to gauge the cooking process for multiple patties at a time, and cameras allowed the robot to ‘see’ its surroundings.

A system that digitally sent tickets to the kitchen from the restaurant’s front counter kept Flippy on top of how many burgers it should be cooking at any given time. Its key tasks were pulling raw patties from a stack and placing them on the grill, tracking each burger’s cook time and temperature, and transferring cooked burgers to a plate.

The new and improved Flippy can do all this and more. It can cook 19 different foods, including chicken wings, onion rings, french fries, and even the Impossible Burger (which, as you may know, isn’t actually made of meat, and that means it’s a little trickier to grill it to perfection).

Flippy’s handiwork. Image Credit: Miso Robotics
And instead of its body sitting on a cart on wheels (which took up a lot of space and meant the robot’s arm could get in the way of human employees), it’s now attached to a rail along the stove’s hood, and can move along the rail to access both the grill and the fryer (provided they’re next to each other, which in many fast food restaurants they are). In fact, Flippy has a new acronym attached to its name: ROAR, which stands for Robot on a Rail.

Flippy ROAR in action, artist rendering. Image Credit: Miso Robotics
Sensors equipped with laser make it safer for human employees to work near Flippy. The bot can automatically switch between different tools, such as a spatula for flipping patties and tongs for gripping the handle of a fryer basket. Its AI software will enable it to learn new skills over time.

Flippy’s interface. Image Credit: Miso Robotics
The first big restaurant chain to go all-in on Flippy was White Castle, which in July announced plans to pilot Flippy ROAR before year’s end. And just last month, Miso made the bot commercially available. The current cost is $30,000 (plus a monthly fee of $1,500 for use of the software), but the company hopes to bring the price down to $20,000 within the next year.

According to Business Insider, demand for the fast food robot is through the roof, probably given a significant boost by the pandemic—thanks, Covid-19. The pace of automation has picked up across multiple sectors, and will likely continue to accelerate as companies look to insure themselves against additional losses.

So for the immediate future, it seems that no matter what happens, we don’t have to worry about the supply of burgers, fries, onion rings, chicken wings, and the like running out.

Now if only Flippy had a cousin—perhaps named Leafy—who could chop vegetables and greens and put together fresh-made salads…

Maybe that can be Miso Robotics’ next project.

Image Credit: Miso Robotics Continue reading

Posted in Human Robots

#437491 3.2 Billion Images and 720,000 Hours of ...

Twitter over the weekend “tagged” as manipulated a video showing US Democratic presidential candidate Joe Biden supposedly forgetting which state he’s in while addressing a crowd.

Biden’s “hello Minnesota” greeting contrasted with prominent signage reading “Tampa, Florida” and “Text FL to 30330.”

The Associated Press’s fact check confirmed the signs were added digitally and the original footage was indeed from a Minnesota rally. But by the time the misleading video was removed it already had more than one million views, The Guardian reports.

A FALSE video claiming Biden forgot what state he was in was viewed more than 1 million times on Twitter in the past 24 hours

In the video, Biden says “Hello, Minnesota.”

The event did indeed happen in MN — signs on stage read MN

But false video edited signs to read Florida pic.twitter.com/LdHQVaky8v

— Donie O'Sullivan (@donie) November 1, 2020

If you use social media, the chances are you see (and forward) some of the more than 3.2 billion images and 720,000 hours of video shared daily. When faced with such a glut of content, how can we know what’s real and what’s not?

While one part of the solution is an increased use of content verification tools, it’s equally important we all boost our digital media literacy. Ultimately, one of the best lines of defense—and the only one you can control—is you.

Seeing Shouldn’t Always Be Believing
Misinformation (when you accidentally share false content) and disinformation (when you intentionally share it) in any medium can erode trust in civil institutions such as news organizations, coalitions and social movements. However, fake photos and videos are often the most potent.

For those with a vested political interest, creating, sharing and/or editing false images can distract, confuse and manipulate viewers to sow discord and uncertainty (especially in already polarized environments). Posters and platforms can also make money from the sharing of fake, sensationalist content.

Only 11-25 percent of journalists globally use social media content verification tools, according to the International Centre for Journalists.

Could You Spot a Doctored Image?
Consider this photo of Martin Luther King Jr.

Dr. Martin Luther King Jr. Giving the middle finger #DopeHistoricPics pic.twitter.com/5W38DRaLHr

— Dope Historic Pics (@dopehistoricpic) December 20, 2013

This altered image clones part of the background over King Jr’s finger, so it looks like he’s flipping off the camera. It has been shared as genuine on Twitter, Reddit, and white supremacist websites.

In the original 1964 photo, King flashed the “V for victory” sign after learning the US Senate had passed the civil rights bill.

“Those who love peace must learn to organize as effectively as those who love war.”
Dr. Martin Luther King Jr.

This photo was taken on June 19th, 1964, showing Dr King giving a peace sign after hearing that the civil rights bill had passed the senate. @snopes pic.twitter.com/LXHmwMYZS5

— Willie's Reserve (@WilliesReserve) January 21, 2019

Beyond adding or removing elements, there’s a whole category of photo manipulation in which images are fused together.

Earlier this year, a photo of an armed man was photoshopped by Fox News, which overlaid the man onto other scenes without disclosing the edits, the Seattle Times reported.

You mean this guy who’s been photoshopped into three separate photos released by Fox News? pic.twitter.com/fAXpIKu77a

— Zander Yates ザンダーイェーツ (@ZanderYates) June 13, 2020

Similarly, the image below was shared thousands of times on social media in January, during Australia’s Black Summer bushfires. The AFP’s fact check confirmed it is not authentic and is actually a combination of several separate photos.

Image is more powerful than screams of Greta. A silent girl is holding a koala. She looks straight at you from the waters of the ocean where they found a refuge. She is wearing a breathing mask. A wall of fire is behind them. I do not know the name of the photographer #Australia pic.twitter.com/CrTX3lltdh

— EVC Music (@EVCMusicUK) January 6, 2020

Fully and Partially Synthetic Content
Online, you’ll also find sophisticated “deepfake” videos showing (usually famous) people saying or doing things they never did. Less advanced versions can be created using apps such as Zao and Reface.

Or, if you don’t want to use your photo for a profile picture, you can default to one of several websites offering hundreds of thousands of AI-generated, photorealistic images of people.

These people don’t exist, they’re just images generated by artificial intelligence. Generated Photos, CC BY

Editing Pixel Values and the (not so) Simple Crop
Cropping can greatly alter the context of a photo, too.

We saw this in 2017, when a US government employee edited official pictures of Donald Trump’s inauguration to make the crowd appear bigger, according to The Guardian. The staffer cropped out the empty space “where the crowd ended” for a set of pictures for Trump.

Views of the crowds at the inaugurations of former US President Barack Obama in 2009 (left) and President Donald Trump in 2017 (right). AP

But what about edits that only alter pixel values such as color, saturation, or contrast?

One historical example illustrates the consequences of this. In 1994, Time magazine’s cover of OJ Simpson considerably “darkened” Simpson in his police mugshot. This added fuel to a case already plagued by racial tension, to which the magazine responded, “No racial implication was intended, by Time or by the artist.”

Tools for Debunking Digital Fakery
For those of us who don’t want to be duped by visual mis/disinformation, there are tools available—although each comes with its own limitations (something we discuss in our recent paper).

Invisible digital watermarking has been proposed as a solution. However, it isn’t widespread and requires buy-in from both content publishers and distributors.

Reverse image search (such as Google’s) is often free and can be helpful for identifying earlier, potentially more authentic copies of images online. That said, it’s not foolproof because it:

Relies on unedited copies of the media already being online.
Doesn’t search the entire web.
Doesn’t always allow filtering by publication time. Some reverse image search services such as TinEye support this function, but Google’s doesn’t.
Returns only exact matches or near-matches, so it’s not thorough. For instance, editing an image and then flipping its orientation can fool Google into thinking it’s an entirely different one.

Most Reliable Tools Are Sophisticated
Meanwhile, manual forensic detection methods for visual mis/disinformation focus mostly on edits visible to the naked eye, or rely on examining features that aren’t included in every image (such as shadows). They’re also time-consuming, expensive, and need specialized expertise.

Still, you can access work in this field by visiting sites such as Snopes.com—which has a growing repository of “fauxtography.”

Computer vision and machine learning also offer relatively advanced detection capabilities for images and videos. But they too require technical expertise to operate and understand.

Moreover, improving them involves using large volumes of “training data,” but the image repositories used for this usually don’t contain the real-world images seen in the news.

If you use an image verification tool such as the REVEAL project’s image verification assistant, you might need an expert to help interpret the results.

The good news, however, is that before turning to any of the above tools, there are some simple questions you can ask yourself to potentially figure out whether a photo or video on social media is fake. Think:

Was it originally made for social media?
How widely and for how long was it circulated?
What responses did it receive?
Who were the intended audiences?

Quite often, the logical conclusions drawn from the answers will be enough to weed out inauthentic visuals. You can access the full list of questions, put together by Manchester Metropolitan University experts, here.

This article is republished from The Conversation under a Creative Commons license. Read the original article.

Image Credit: Simon Steinberger from Pixabay Continue reading

Posted in Human Robots

#437466 How Future AI Could Recognize a Kangaroo ...

AI is continuously taking on new challenges, from detecting deepfakes (which, incidentally, are also made using AI) to winning at poker to giving synthetic biology experiments a boost. These impressive feats result partly from the huge datasets the systems are trained on. That training is costly and time-consuming, and it yields AIs that can really only do one thing well.

For example, to train an AI to differentiate between a picture of a dog and one of a cat, it’s fed thousands—if not millions—of labeled images of dogs and cats. A child, on the other hand, can see a dog or cat just once or twice and remember which is which. How can we make AIs learn more like children do?

A team at the University of Waterloo in Ontario has an answer: change the way AIs are trained.

Here’s the thing about the datasets normally used to train AI—besides being huge, they’re highly specific. A picture of a dog can only be a picture of a dog, right? But what about a really small dog with a long-ish tail? That sort of dog, while still being a dog, looks more like a cat than, say, a fully-grown Golden Retriever.

It’s this concept that the Waterloo team’s methodology is based on. They described their work in a paper published on the pre-print (or non-peer-reviewed) server arXiv last month. Teaching an AI system to identify a new class of objects using just one example is what they call “one-shot learning.” But they take it a step further, focusing on “less than one shot learning,” or LO-shot learning for short.

LO-shot learning consists of a system learning to classify various categories based on a number of examples that’s smaller than the number of categories. That’s not the most straightforward concept to wrap your head around, so let’s go back to the dogs and cats example. Say you want to teach an AI to identify dogs, cats, and kangaroos. How could that possibly be done without several clear examples of each animal?

The key, the Waterloo team says, is in what they call soft labels. Unlike hard labels, which label a data point as belonging to one specific class, soft labels tease out the relationship or degree of similarity between that data point and multiple classes. In the case of an AI trained on only dogs and cats, a third class of objects, say, kangaroos, might be described as 60 percent like a dog and 40 percent like a cat (I know—kangaroos probably aren’t the best animal to have thrown in as a third category).

“Soft labels can be used to represent training sets using fewer prototypes than there are classes, achieving large increases in sample efficiency over regular (hard-label) prototypes,” the paper says. Translation? Tell an AI a kangaroo is some fraction cat and some fraction dog—both of which it’s seen and knows well—and it’ll be able to identify a kangaroo without ever having seen one.

If the soft labels are nuanced enough, you could theoretically teach an AI to identify a large number of categories based on a much smaller number of training examples.

The paper’s authors use a simple machine learning algorithm called k-nearest neighbors (kNN) to explore this idea more in depth. The algorithm operates under the assumption that similar things are most likely to exist near each other; if you go to a dog park, there will be lots of dogs but no cats or kangaroos. Go to the Australian grasslands and there’ll be kangaroos but no cats or dogs. And so on.

To train a kNN algorithm to differentiate between categories, you choose specific features to represent each category (i.e. for animals you could use weight or size as a feature). With one feature on the x-axis and the other on the y-axis, the algorithm creates a graph where data points that are similar to each other are clustered near each other. A line down the center divides the categories, and it’s pretty straightforward for the algorithm to discern which side of the line new data points should fall on.

The Waterloo team kept it simple and used plots of color on a 2D graph. Using the colors and their locations on the graphs, the team created synthetic data sets and accompanying soft labels. One of the more simplistic graphs is pictured below, along with soft labels in the form of pie charts.

Image Credit: Ilia Sucholutsky & Matthias Schonlau
When the team had the algorithm plot the boundary lines of the different colors based on these soft labels, it was able to split the plot up into more colors than the number of data points it was given in the soft labels.

While the results are encouraging, the team acknowledges that they’re just the first step, and there’s much more exploration of this concept yet to be done. The kNN algorithm is one of the least complex models out there; what might happen when LO-shot learning is applied to a far more complex algorithm? Also, to apply it, you still need to distill a larger dataset down into soft labels.

One idea the team is already working on is having other algorithms generate the soft labels for the algorithm that’s going to be trained using LO-shot; manually designing soft labels won’t always be as easy as splitting up some pie charts into different colors.

LO-shot’s potential for reducing the amount of training data needed to yield working AI systems is promising. Besides reducing the cost and the time required to train new models, the method could also make AI more accessible to industries, companies, or individuals who don’t have access to large datasets—an important step for democratization of AI.

Image Credit: pen_ash from Pixabay Continue reading

Posted in Human Robots

#437460 This Week’s Awesome Tech Stories From ...

ARTIFICIAL INTELLIGENCE
A Radical New Technique Lets AI Learn With Practically No Data
Karen Hao | MIT Technology Review
“Shown photos of a horse and a rhino, and told a unicorn is something in between, [children] can recognize the mythical creature in a picture book the first time they see it. …Now a new paper from the University of Waterloo in Ontario suggests that AI models should also be able to do this—a process the researchers call ‘less than one’-shot, or LO-shot, learning.”

FUTURE
Artificial General Intelligence: Are We Close, and Does It Even Make Sense to Try?
Will Douglas Heaven | MIT Technology Review
“A machine that could think like a person has been the guiding vision of AI research since the earliest days—and remains its most divisive idea. …So why is AGI controversial? Why does it matter? And is it a reckless, misleading dream—or the ultimate goal?”

HEALTH
The Race for a Super-Antibody Against the Coronavirus
Apoorva Mandavilli | The New York Times
“Dozens of companies and academic groups are racing to develop antibody therapies. …But some scientists are betting on a dark horse: Prometheus, a ragtag group of scientists who are months behind in the competition—and yet may ultimately deliver the most powerful antibody.”

SPACE
How to Build a Spacecraft to Save the World
Daniel Oberhaus | Wired
“The goal of the Double Asteroid Redirection Test, or DART, is to slam the [spacecraft] into a small asteroid orbiting a larger asteroid 7 million miles from Earth. …It should be able to change the asteroid’s orbit just enough to be detectable from Earth, demonstrating that this kind of strike could nudge an oncoming threat out of Earth’s way. Beyond that, everything is just an educated guess, which is exactly why NASA needs to punch an asteroid with a robot.”

TRANSPORTATION
Inside Gravity’s Daring Mission to Make Jetpacks a Reality
Oliver Franklin-Wallis | Wired
“The first time someone flies a jetpack, a curious thing happens: just as their body leaves the ground, their legs start to flail. …It’s as if the vestibular system can’t quite believe what’s happening. This isn’t natural. Then suddenly, thrust exceeds weight, and—they’re aloft. …It’s that moment, lift-off, that has given jetpacks an enduring appeal for over a century.”

FUTURE OF FOOD
Inside Singapore’s Huge Bet on Vertical Farming
Megan Tatum | MIT Technology Review
“…to cram all [of Singapore’s] gleaming towers and nearly 6 million people into a land mass half the size of Los Angeles, it has sacrificed many things, including food production. Farms make up no more than 1% of its total land (in the United States it’s 40%), forcing the small city-state to shell out around $10 billion each year importing 90% of its food. Here was an example of technology that could change all that.”

COMPUTING
The Effort to Build the Mathematical Library of the Future
Kevin Hartnett | Quanta
“Digitizing mathematics is a longtime dream. The expected benefits range from the mundane—computers grading students’ homework—to the transcendent: using artificial intelligence to discover new mathematics and find new solutions to old problems.”

Image credit: Kevin Mueller / Unsplash Continue reading

Posted in Human Robots