Department Publications

Image generation from a single image using generative adversarial networks is quite interesting due to the realism of generated images. However, recent approaches need improvement for such realistic and diverse image generation, when the global context of the image is important such as in face, animal, and architectural image generation. This is mainly due to the use of fewer convolutional layers for mainly capturing the patch statistics and, thereby, not being able to capture global statistics very well. We solve this problem by using attention blocks at selected scales and feeding a random Gaussian blurred image to the discriminator for training. Our results are visually better than the state-of-the-art particularly in generating images that require global context. The diversity of our image generation, measured using the average standard deviation of pixels, is also better.

This paper introduces a novel locomotion strategy of combining wheeled and legged actuation in a parallel structure. This enables the robot to traverse rough terrains using the articulated limbs and use wheeled locomotion for more efficient traversal over smooth terrains. Authors present a prototype robot architecture to embody the concept and show experimentally the robot traversing a variety of indoor and outdoor terrains. Additionally, a novel butterfly gait is introduced for the robot that enables stable and efficient traversal of challenging terrains.