While great strides have been made in detecting and localizing specific objects in natural images, the bottom-up segmentation of unknown, generic objects remains a difficult chall...
Abstract. Most previous methods for generic object recognition explicitly or implicitly assume that an image contains objects from a single category, although objects from multiple...
Takahiro Okabe, Yuhi Kondo, Kris M. Kitani, Yoichi...
Our group has previously used machine learning techniques to develop computational systems to automatically analyse fluorescence microscope images and classify the location of the ...
The need to register data is abundant in applications such as: world modeling, part inspection and manufacturing, object recognition, pose estimation, robotic navigation, and reve...
— Opportunistic spectrum access and adaptive power management are effective techniques to improve throughput, delay performance, and energy efficiency for wireless networks. In ...