With the great capabilities of deep classifiers for radar data processing come the risks of learning dataset-specific features that do not generalize well. In this work, the robustness of two deep convolutional architectures, trained and tested on the same data, is evaluated. When standard training practice is followed, both classifiers exhibit sensitivity to subtle temporal shifts of the input representation, an augmentation that carries minimal semantic content. Furthermore, the models are extremely susceptible to adversarial examples. Both small temporal shifts and adversarial examples are a result of a model overfitting on features that do not generalize well. As a remedy, it is shown that training on adversarial examples and temporally augmented samples can reduce this effect and lead to models that generalise better. Finally, models operating on cadence-velocity diagram representation rather than Doppler-time are demonstrated to be naturally more immune to adversarial examples.

Those that climb need to be both fast and stable to avoid predation and find food. A robot made to mimic their movements has now shown how the rotation of their legs and the speed with which they move up vertical surfaces helps them climb efficiently. "Most lizards look a lot like other lizards," says Christofer Clemente at University of the Sunshine Coast, Australia. To find out why, Clemente and his team built a robot based on a lizard's body plan to explore its efficiency. It is about 24 centimetres long, and its legs and feet were programmed to mimic the gait of climbing lizards.

The study, which featured a University of Queensland researcher, used a slow motion camera to capture the nuanced movement of eight species of Australian agamid lizards that run on two legs -- an action known as'bipedal' movement. UQ School of Biological Sciences researcher Nicholas Wu said the study's findings challenged existing mathematical models based on the animals' movement. "There was an existing understanding that the backwards shift in these lizards' centre of mass, combined with quick bursts of acceleration, caused them to start running on two legs at a certain point," he said. "What we found though is that some lizards run bipedally sooner than expected, by moving their body back and winging their tail up. "This means that they could run bipedally for longer, perhaps to overcome obstacles in their path." Lead author Christofer Clemente from the University of the Sunshine Coast said these results may have important implications for the design of bio-inspired robotic devices. "We're still teasing out why these species have evolved to run like this in the first place, but as we learn more, it's clear that these lessons from nature may be able to be integrated into robotics," Dr. Clemente said. "It's been suggested that this movement might have something to do with increasing vision in moments of urgency, by elevating the head at the same time and helping to navigate over obstacles.