Detecting key frames in videos is a common problem in many applications such as video classification, action recognition and video summarization. These tasks can be performed more efficiently using only a handful of key frames rather than the full video. Existing key frame detection approaches are mostly designed for supervised learning and require manual labelling of key frames in a large corpus of training data to train the models. Labelling requires human annotators from different backgrounds to annotate key frames in videos which is not only expensive and time consuming but also prone to subjective errors and inconsistencies between the labelers. To overcome these problems, we propose an automatic self-supervised method for detecting key frames in a video. Our method comprises a two-stream ConvNet and a novel automatic annotation architecture able to reliably annotate key frames in a video for self-supervised learning of the ConvNet. The proposed ConvNet learns deep appearance and motion features to detect frames that are unique. The trained network is then able to detect key frames in test videos. Extensive experiments on UCF101 human action and video summarization VSUMM datasets demonstrates the effectiveness of our proposed method.

No abstract

No abstract

The presence of stripe nonuniformity severely degrades the image quality and affects the performance in many infrared (IR) sensing applications. Prior works correct the nonuniformity by using similar spatial representations, which inevitably damage some detailed structures of the image. In this paper, we instead take advantage of spectral prior of stripe noise to solve its correction problem in single IR image. We first analyse the significant spectral difference between stripes and image structures and utilize this knowledge to characterize stripe nonuniformity. Then a two-stage filtering strategy is adopted combining spectral and spatial filtering. The proposed method enables stripe nonuniformity to be eliminated from coarse to fine, thus preserving image details well. Extensive experiments on simulated images and raw IR images demonstrate that the proposed method achieves superior correction performance over the recent state-of-the-art methods.

No abstract