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

Since manually labeling training data is slow and expensive, recent industrial and scientific research efforts have turned to weaker or noisier forms of supervision sources. However, existing weak supervision approaches fail to model multi-resolution sources for sequential data, like video, that can assign labels to individual elements or collections of elements in a sequence. A key challenge in weak supervision is estimating the unknown accuracies and correlations of these sources without using labeled data. Multi-resolution sources exacerbate this challenge due to complex correlations and sample complexity that scales in the length of the sequence. We propose Dugong, the first framework to model multi-resolution weak supervision sources with complex correlations to assign probabilistic labels to training data. Theoretically, we prove that Dugong, under mild conditions, can uniquely recover the unobserved accuracy and correlation parameters and use parameter sharing to improve sample complexity. Our method assigns clinician-validated labels to population-scale biomedical video repositories, helping outperform traditional supervision by 36.8 F1 points and addressing a key use case where machine learning has been severely limited by the lack of expert labeled data. On average, Dugong improves over traditional supervision by 16.0 F1 points and existing weak supervision approaches by 24.2 F1 points across several video and sensor classification tasks.

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

Since manually labeling training data is slow and expensive, recent industrial and scientific research efforts have turned to weaker or noisier forms of supervision sources. However, existing weak supervision approaches fail to model multi-resolution sources for sequential data, like video, that can assign labels to individual elements or collections of elements in a sequence. A key challenge in weak supervision is estimating the unknown accuracies and correlations of these sources without using labeled data. Multi-resolution sources exacerbate this challenge due to complex correlations and sample complexity that scales in the length of the sequence. We propose Dugong, the first framework to model multi-resolution weak supervision sources with complex correlations to assign probabilistic labels to training data.

Since manually labeling training data is slow and expensive, recent industrial and scientific research efforts have turned to weaker or noisier forms of supervision sources. However, existing weak supervision approaches fail to model multi-resolution sources for sequential data, like video, that can assign labels to individual elements or collections of elements in a sequence. A key challenge in weak supervision is estimating the unknown accuracies and correlations of these sources without using labeled data. Multi-resolution sources exacerbate this challenge due to complex correlations and sample complexity that scales in the length of the sequence. We propose Dugong, the first framework to model multi-resolution weak supervision sources with complex correlations to assign probabilistic labels to training data.

Since manually labeling training data is slow and expensive, recent industrial and scientific research efforts have turned to weaker or noisier forms of supervision sources. However, existing weak supervision approaches fail to model multi-resolution sources for sequential data, like video, that can assign labels to individual elements or collections of elements in a sequence. A key challenge in weak supervision is estimating the unknown accuracies and correlations of these sources without using labeled data. Multi-resolution sources exacerbate this challenge due to complex correlations and sample complexity that scales in the length of the sequence. We propose Dugong, the first framework to model multi-resolution weak supervision sources with complex correlations to assign probabilistic labels to training data.

It's obviously important to Australians to make sure their koala population is closely tracked -- but how can you do so when the suckers live in forests and climb trees all the time? A new project from Queensland University of Technology combines some well-known techniques in a new way to help keep an eye on wild populations of the famous and soft marsupials. They used a drone equipped with a heat-sensing camera, then ran the footage through a deep learning model trained to look for koala-like heat signatures. It's similar in some ways to an earlier project from QUT in which dugongs -- endangered sea cows -- were counted along the shore via aerial imagery and machine learning. But this is considerably harder.

On the west coast of Australia, Amanda Hodgson is launching drones out towards the Indian Ocean so that they can photograph the water from above. The photos are a way of locating dugongs, or sea cows, in the bay near Perth--part of an effort to prevent the extinction of these endangered marine mammals. The trouble is that Hodgson and her team don't have the time needed to examine all those aerial photos. There are too many of them--about 45,000--and spotting the dugongs is far too difficult for the untrained eye. Deep learning is remaking Google, Facebook, Microsoft, and Amazon.

On the west coast of Australia, Amanda Hodgson is launching drones out towards the Indian Ocean so that they can photograph the water from above. The photos are a way of locating dugongs, or sea cows, in the bay near Perth--part of an effort to prevent the extinction of these endangered marine mammals. The trouble is that Hodgson and her team don't have the time needed to examine all those aerial photos. There are too many of them--about 45,000--and spotting the dugongs is far too difficult for the untrained eye. Deep learning is remaking Google, Facebook, Microsoft, and Amazon.

Dugong expert Dr Amanda Hodgson estimates she has stared at more than 30,000 photographs of blue water. "It's really taxing on your eyes and it's hard to maintain concentration." The researcher from WA's Murdoch University has been scanning pictures captured by aerial drones in a search for dugongs, to work out their population, size and location. Globally dugongs are classed as "vulnerable to extinction" and are found in waters off the northern half of Australia. "There are areas where they're quite vulnerable because their habitat overlaps with coastal development."