Lifelogging is the act of recording some aspect of your life in digital format. A basic and common form of lifelogging is the creation and maintenance of blogs, which are typically textual in nature, though often with multimedia elements. In this paper we are concerned with visual lifelogging, a new form of lifelogging based on the passive capture of photos of a person’s experiences. We examine...

We are describing collaborative authoring technology over the network. One of the components in our authoring technology is photo recommendation system, in which recommended photos will be used in collaborative video authoring and social album. The recommended photos are derived from Facebook, by referring properties of photos currently available on the authoring library. The photo properties i...

Digital cameras have made it much easier to take photos, but organizing those photos is difficult. As a result, many people have thousands of photos in some miscellaneous folder on their hard disk . If computer can understand and manage these photos for us, we can save time. Also it will be useful for indexing and searching the web images. In this paper we propose an image annotation system wit...

CNNs have massively improved performance in object detection in photographs. However research into object detection in artwork remains limited. We show state-of-the-art performance on a challenging dataset, People-Art, which contains people from photos, cartoons and 41 different artwork movements. We achieve this high performance by fine-tuning a CNN for this task, thus also demonstrating that ...

We are exploring the use of collaborative games to generate meaningful textual tags for photos. We have designed PhotoPlay to take advantage of the social engagement typical of board games and provide a collocated ludic environment conducive to the creation of text tags. We evaluated PhotoPlay and found that it was fun and socially engaging for players. The milieu of the game also facilitated p...

The present study aims to replicate and extend Rule and Ambady (2010a)'s findings that Republicans and Democrats could be differentiated by face. In Experiment 1, undergraduates categorized 50 gray-scale full-face photos of candidates of the two major political parties in Taiwan, the Kuomingtang (KMT) and the Democratic Progressive Party (DPP). Using identical stimuli and procedure, Experiment ...

In recent years, massive amounts of identified and unidentified facial data have become available—often publicly so—through Web 2.0 applications. So have also the infrastructure and technologies necessary to navigate through those data in real time, matching individuals across online services, independently of their knowledge or consent. In the literature on statistical re-identification [5, 6]...