Neural Information Processing Systems http://nips.cc/

Neural Information Processing Systems http://nips.cc/

Neural Information Processing Systems http://nips.cc/

Reasoning is an important ability that we learn from a very early age.

How many atoms are there in the observable universe? Current estimates point to a number we would write as 1 followed by 80 zeroes, or 1080. If you peered inside each of these atoms and counted their subatomic particles, you could count a bit higher. But what happens beyond that? Take 1090 – even if you counted every atom and subatomic particle in the known universe, you wouldn't reach this number. In some sense, 1090 has no relation to physical reality.

We introduce a flexible framework that produces high-quality almost-exact matches for causal inference. Most prior work in matching uses ad-hoc distance metrics, often leading to poor quality matches, particularly when there are irrelevant covariates. In this work, we learn an interpretable distance metric for matching, which leads to substantially higher quality matches. The learned distance metric stretches the covariate space according to each covariate's contribution to outcome prediction: this stretching means that mismatches on important covariates carry a larger penalty than mismatches on irrelevant covariates. Our ability to learn flexible distance metrics leads to matches that are interpretable and useful for the estimation of conditional average treatment effects.

Visual question answering (VQA) is challenging not only because the model has to handle multi-modal information, but also because it is just so hard to collect sufficient training examples -- there are too many questions one can ask about an image. As a result, a VQA model trained solely on human-annotated examples could easily over-fit specific question styles or image contents that are being asked, leaving the model largely ignorant about the sheer diversity of questions. Existing methods address this issue primarily by introducing an auxiliary task such as visual grounding, cycle consistency, or debiasing. In this paper, we take a drastically different approach. We found that many of the "unknowns" to the learned VQA model are indeed "known" in the dataset implicitly. For instance, questions asking about the same object in different images are likely paraphrases; the number of detected or annotated objects in an image already provides the answer to the "how many" question, even if the question has not been annotated for that image. Building upon these insights, we present a simple data augmentation pipeline SimpleAug to turn this "known" knowledge into training examples for VQA. We show that these augmented examples can notably improve the learned VQA models' performance, not only on the VQA-CP dataset with language prior shifts but also on the VQA v2 dataset without such shifts. Our method further opens up the door to leverage weakly-labeled or unlabeled images in a principled way to enhance VQA models. Our code and data are publicly available at https://github.com/heendung/simpleAUG.

With accolades showered upon them, seemingly perfect educational pedigrees, and conversations focused mostly on their groundbreaking work, it's hard to remember that artificial intelligence (AI) researchers are real people. They're people who take ballet classes as an adult, who suffer from anxiety, and love art. They're also people who feel giddy over fresh flowers, have a difficult time staying organized, and roll out of bed just in time for their first meeting. Devi Parikh, an associate professor in the Machine Learning Center at Georgia Tech (ML@GT) and School of Interactive Computing (IC,) is working to change that with her new podcast and video series, Humans of AI: Stories, Not Stats. The series, which launches on Oct. 20, features 18 conversations with leading AI researchers, including Jeff Dean (head of AI at Google), Animashree Anandkumar (Bren Professor at California Institute of Technology and Director of Machine Learning Research at NVIDIA), Ayanna Howard (School of IC chair and professor at Georgia Tech),and Timnit Gebru (co-lead of ethical AI at Google.)

Fox News Flash top headlines are here. Check out what's clicking on Foxnews.com. A robot dog armed with a sniper rifle was unveiled this week in Washington, D.C. at the annual meeting of the Association of the United States Army. The robot, developed by Ghost Robotics, carries a SWORD Defense Systems Special Purpose Unmanned Rifle (SPUR). Check out the latest partner payloads @AUSAorg Wash DC.

The US military may be getting a dog-like quadruped robot armed with a sniper rifle. The robot, developed by Ghost Robotics of Philadelphia, is a new version of its Vision series of legged robots. The US Air Force is currently testing an unarmed version of these robots for use as perimeter security at the Tyndall Air Force Base in Florida. Ghost Robotics displayed the armed version at the annual meeting of the Association of the United States Army held in Washington DC this week. The robot is fitted with a Special Purpose Unmanned Rifle pod from Sword Defense, with a powerful 6.5mm sniper rifle.