Tag Archives: ucla
#434559 Can AI Tell the Difference Between a ...
Scarcely a day goes by without another headline about neural networks: some new task that deep learning algorithms can excel at, approaching or even surpassing human competence. As the application of this approach to computer vision has continued to improve, with algorithms capable of specialized recognition tasks like those found in medicine, the software is getting closer to widespread commercial use—for example, in self-driving cars. Our ability to recognize patterns is a huge part of human intelligence: if this can be done faster by machines, the consequences will be profound.
Yet, as ever with algorithms, there are deep concerns about their reliability, especially when we don’t know precisely how they work. State-of-the-art neural networks will confidently—and incorrectly—classify images that look like television static or abstract art as real-world objects like school-buses or armadillos. Specific algorithms could be targeted by “adversarial examples,” where adding an imperceptible amount of noise to an image can cause an algorithm to completely mistake one object for another. Machine learning experts enjoy constructing these images to trick advanced software, but if a self-driving car could be fooled by a few stickers, it might not be so fun for the passengers.
These difficulties are hard to smooth out in large part because we don’t have a great intuition for how these neural networks “see” and “recognize” objects. The main insight analyzing a trained network itself can give us is a series of statistical weights, associating certain groups of points with certain objects: this can be very difficult to interpret.
Now, new research from UCLA, published in the journal PLOS Computational Biology, is testing neural networks to understand the limits of their vision and the differences between computer vision and human vision. Nicholas Baker, Hongjing Lu, and Philip J. Kellman of UCLA, alongside Gennady Erlikhman of the University of Nevada, tested a deep convolutional neural network called VGG-19. This is state-of-the-art technology that is already outperforming humans on standardized tests like the ImageNet Large Scale Visual Recognition Challenge.
They found that, while humans tend to classify objects based on their overall (global) shape, deep neural networks are far more sensitive to the textures of objects, including local color gradients and the distribution of points on the object. This result helps explain why neural networks in image recognition make mistakes that no human ever would—and could allow for better designs in the future.
In the first experiment, a neural network was trained to sort images into 1 of 1,000 different categories. It was then presented with silhouettes of these images: all of the local information was lost, while only the outline of the object remained. Ordinarily, the trained neural net was capable of recognizing these objects, assigning more than 90% probability to the correct classification. Studying silhouettes, this dropped to 10%. While human observers could nearly always produce correct shape labels, the neural networks appeared almost insensitive to the overall shape of the images. On average, the correct object was ranked as the 209th most likely solution by the neural network, even though the overall shapes were an exact match.
A particularly striking example arose when they tried to get the neural networks to classify glass figurines of objects they could already recognize. While you or I might find it easy to identify a glass model of an otter or a polar bear, the neural network classified them as “oxygen mask” and “can opener” respectively. By presenting glass figurines, where the texture information that neural networks relied on for classifying objects is lost, the neural network was unable to recognize the objects by shape alone. The neural network was similarly hopeless at classifying objects based on drawings of their outline.
If you got one of these right, you’re better than state-of-the-art image recognition software. Image Credit: Nicholas Baker, Hongjing Lu, Gennady Erlikhman, Philip J. Kelman. “Deep convolutional networks do not classify based on global object shape.” Plos Computational Biology. 12/7/18. / CC BY 4.0
When the neural network was explicitly trained to recognize object silhouettes—given no information in the training data aside from the object outlines—the researchers found that slight distortions or “ripples” to the contour of the image were again enough to fool the AI, while humans paid them no mind.
The fact that neural networks seem to be insensitive to the overall shape of an object—relying instead on statistical similarities between local distributions of points—suggests a further experiment. What if you scrambled the images so that the overall shape was lost but local features were preserved? It turns out that the neural networks are far better and faster at recognizing scrambled versions of objects than outlines, even when humans struggle. Students could classify only 37% of the scrambled objects, while the neural network succeeded 83% of the time.
Humans vastly outperform machines at classifying object (a) as a bear, while the machine learning algorithm has few problems classifying the bear in figure (b). Image Credit: Nicholas Baker, Hongjing Lu, Gennady Erlikhman, Philip J. Kelman. “Deep convolutional networks do not classify based on global object shape.” Plos Computational Biology. 12/7/18. / CC BY 4.0
“This study shows these systems get the right answer in the images they were trained on without considering shape,” Kellman said. “For humans, overall shape is primary for object recognition, and identifying images by overall shape doesn’t seem to be in these deep learning systems at all.”
Naively, one might expect that—as the many layers of a neural network are modeled on connections between neurons in the brain and resemble the visual cortex specifically—the way computer vision operates must necessarily be similar to human vision. But this kind of research shows that, while the fundamental architecture might resemble that of the human brain, the resulting “mind” operates very differently.
Researchers can, increasingly, observe how the “neurons” in neural networks light up when exposed to stimuli and compare it to how biological systems respond to the same stimuli. Perhaps someday it might be possible to use these comparisons to understand how neural networks are “thinking” and how those responses differ from humans.
But, as yet, it takes a more experimental psychology to probe how neural networks and artificial intelligence algorithms perceive the world. The tests employed against the neural network are closer to how scientists might try to understand the senses of an animal or the developing brain of a young child rather than a piece of software.
By combining this experimental psychology with new neural network designs or error-correction techniques, it may be possible to make them even more reliable. Yet this research illustrates just how much we still don’t understand about the algorithms we’re creating and using: how they tick, how they make decisions, and how they’re different from us. As they play an ever-greater role in society, understanding the psychology of neural networks will be crucial if we want to use them wisely and effectively—and not end up missing the woods for the trees.
Image Credit: Irvan Pratama / Shutterstock.com Continue reading
Posted in Human Robots
#432236 Why Hasn’t AI Mastered Language ...
In the myth about the Tower of Babel, people conspired to build a city and tower that would reach heaven. Their creator observed, “And now nothing will be restrained from them, which they have imagined to do.” According to the myth, God thwarted this effort by creating diverse languages so that they could no longer collaborate.
In our modern times, we’re experiencing a state of unprecedented connectivity thanks to technology. However, we’re still living under the shadow of the Tower of Babel. Language remains a barrier in business and marketing. Even though technological devices can quickly and easily connect, humans from different parts of the world often can’t.
Translation agencies step in, making presentations, contracts, outsourcing instructions, and advertisements comprehensible to all intended recipients. Some agencies also offer “localization” expertise. For instance, if a company is marketing in Quebec, the advertisements need to be in Québécois French, not European French. Risk-averse companies may be reluctant to invest in these translations. Consequently, these ventures haven’t achieved full market penetration.
Global markets are waiting, but AI-powered language translation isn’t ready yet, despite recent advancements in natural language processing and sentiment analysis. AI still has difficulties processing requests in one language, without the additional complications of translation. In November 2016, Google added a neural network to its translation tool. However, some of its translations are still socially and grammatically odd. I spoke to technologists and a language professor to find out why.
“To Google’s credit, they made a pretty massive improvement that appeared almost overnight. You know, I don’t use it as much. I will say this. Language is hard,” said Michael Housman, chief data science officer at RapportBoost.AI and faculty member of Singularity University.
He explained that the ideal scenario for machine learning and artificial intelligence is something with fixed rules and a clear-cut measure of success or failure. He named chess as an obvious example, and noted machines were able to beat the best human Go player. This happened faster than anyone anticipated because of the game’s very clear rules and limited set of moves.
Housman elaborated, “Language is almost the opposite of that. There aren’t as clearly-cut and defined rules. The conversation can go in an infinite number of different directions. And then of course, you need labeled data. You need to tell the machine to do it right or wrong.”
Housman noted that it’s inherently difficult to assign these informative labels. “Two translators won’t even agree on whether it was translated properly or not,” he said. “Language is kind of the wild west, in terms of data.”
Google’s technology is now able to consider the entirety of a sentence, as opposed to merely translating individual words. Still, the glitches linger. I asked Dr. Jorge Majfud, Associate Professor of Spanish, Latin American Literature, and International Studies at Jacksonville University, to explain why consistently accurate language translation has thus far eluded AI.
He replied, “The problem is that considering the ‘entire’ sentence is still not enough. The same way the meaning of a word depends on the rest of the sentence (more in English than in Spanish), the meaning of a sentence depends on the rest of the paragraph and the rest of the text, as the meaning of a text depends on a larger context called culture, speaker intentions, etc.”
He noted that sarcasm and irony only make sense within this widened context. Similarly, idioms can be problematic for automated translations.
“Google translation is a good tool if you use it as a tool, that is, not to substitute human learning or understanding,” he said, before offering examples of mistranslations that could occur.
“Months ago, I went to buy a drill at Home Depot and I read a sign under a machine: ‘Saw machine.’ Right below it, the Spanish translation: ‘La máquina vió,’ which means, ‘The machine did see it.’ Saw, not as a noun but as a verb in the preterit form,” he explained.
Dr. Majfud warned, “We should be aware of the fragility of their ‘interpretation.’ Because to translate is basically to interpret, not just an idea but a feeling. Human feelings and ideas that only humans can understand—and sometimes not even we, humans, understand other humans.”
He noted that cultures, gender, and even age can pose barriers to this understanding and also contended that an over-reliance on technology is leading to our cultural and political decline. Dr. Majfud mentioned that Argentinean writer Julio Cortázar used to refer to dictionaries as “cemeteries.” He suggested that automatic translators could be called “zombies.”
Erik Cambria is an academic AI researcher and assistant professor at Nanyang Technological University in Singapore. He mostly focuses on natural language processing, which is at the core of AI-powered language translation. Like Dr. Majfud, he sees the complexity and associated risks. “There are so many things that we unconsciously do when we read a piece of text,” he told me. Reading comprehension requires multiple interrelated tasks, which haven’t been accounted for in past attempts to automate translation.
Cambria continued, “The biggest issue with machine translation today is that we tend to go from the syntactic form of a sentence in the input language to the syntactic form of that sentence in the target language. That’s not what we humans do. We first decode the meaning of the sentence in the input language and then we encode that meaning into the target language.”
Additionally, there are cultural risks involved with these translations. Dr. Ramesh Srinivasan, Director of UCLA’s Digital Cultures Lab, said that new technological tools sometimes reflect underlying biases.
“There tend to be two parameters that shape how we design ‘intelligent systems.’ One is the values and you might say biases of those that create the systems. And the second is the world if you will that they learn from,” he told me. “If you build AI systems that reflect the biases of their creators and of the world more largely, you get some, occasionally, spectacular failures.”
Dr. Srinivasan said translation tools should be transparent about their capabilities and limitations. He said, “You know, the idea that a single system can take languages that I believe are very diverse semantically and syntactically from one another and claim to unite them or universalize them, or essentially make them sort of a singular entity, it’s a misnomer, right?”
Mary Cochran, co-founder of Launching Labs Marketing, sees the commercial upside. She mentioned that listings in online marketplaces such as Amazon could potentially be auto-translated and optimized for buyers in other countries.
She said, “I believe that we’re just at the tip of the iceberg, so to speak, with what AI can do with marketing. And with better translation, and more globalization around the world, AI can’t help but lead to exploding markets.”
Image Credit: igor kisselev / Shutterstock.com Continue reading
Posted in Human Robots
#429937 “Walk This Way” – An ...
The NABiRoS (Non-Anthropomorphic Bipedal Robotic System) novel approach to an efficient “walking” system for humanoids places the legs in the sagittal plane and adds compliance to the feet, reducing the actuated degrees of freedom to just four! This drastically reduces … Continue reading
Posted in Human Robots
#428575 Meet BALLU, UCLA’s Humanoid Blimp ...
A hybrid between a blimp and a humanoid robot can do amazing things Continue reading
Posted in Human Robots