Vision and language understanding has emerged as a subject undergoing intense study in Artificial Intelligence. Among many tasks in this line of research, visual question answering (VQA) has been one of the most successful ones, where the goal is to learn a model that understands visual content at region-level details and finds their associations with pairs of questions and answers in the natural language form. Despite the rapid progress in the past few years, most existing work in VQA have focused primarily on images. In this paper, we focus on extending VQA to the video domain and contribute to the literature in three important ways. First, we propose three new tasks designed specifically for video VQA, which require spatio-temporal reasoning from videos to answer questions correctly. Next, we introduce a new large-scale dataset for video VQA named TGIF-QA that extends existing VQA work with our new tasks. Finally, we propose a dual-LSTM based approach with both spatial and temporal attention, and show its effectiveness over conventional VQA techniques through empirical evaluations.

Building upon literature suggesting low Internet use among racial/ethnic minorities and socioeconomically disadvantaged groups, this study examined how race/ethnicity and socioeconomic status (SES) influence the Internet use for health information, addressing both independent and interactive effects. Using data from 17,704 older adults in the California Health Interview Survey, logistic regression models were estimated with race/ethnicity (Whites, African Americans, Latinos, and Asians), SES index, and the interaction between race/ethnicity and SES index. Overall, approximately 40% of participants were Internet-users for health information. Direct effects of race/ethnicity and SES-and their interactions-were all found to be significant. Minority status combined with the lowest levels of SES substantially reduced the odds of using Internet for health information. Findings suggest the combination of racial/ethnic minority status and low SES as a source of digital divide, and provide implications for Internet technology training for the target population.

The findings support the importance of psychosocial factors in modifying the association between disability and depression and suggest that efforts to enhance positive psychosocial attributes should be emphasized in interventions for older adults.

No abstract