Video Stream Retrieval of Unseen Queries using Semantic Memory
This addresses the challenge of on-line video stream retrieval for users needing to search live broadcasts with unpredictable queries, though it is incremental in adapting existing methods to this specific domain.
The paper tackles the problem of retrieving live video streams for arbitrary, unseen queries by using semantic relatedness to pre-trained concept classifiers and adapting to shifting content with memory pooling and welling methods. They report results on three large-scale video datasets adapted for stream retrieval tasks, demonstrating efficacy in both new stream retrieval and traditional video tasks.
Retrieval of live, user-broadcast video streams is an under-addressed and increasingly relevant challenge. The on-line nature of the problem requires temporal evaluation and the unforeseeable scope of potential queries motivates an approach which can accommodate arbitrary search queries. To account for the breadth of possible queries, we adopt a no-example approach to query retrieval, which uses a query's semantic relatedness to pre-trained concept classifiers. To adapt to shifting video content, we propose memory pooling and memory welling methods that favor recent information over long past content. We identify two stream retrieval tasks, instantaneous retrieval at any particular time and continuous retrieval over a prolonged duration, and propose means for evaluating them. Three large scale video datasets are adapted to the challenge of stream retrieval. We report results for our search methods on the new stream retrieval tasks, as well as demonstrate their efficacy in a traditional, non-streaming video task.