Abstract
AbstractDoes a query image with much higher resolution than that of the gallery image also affect the pedestrian re‐identification performance? If so, and how does it affect performance? The study proposes a novel framework for performing high‐resolution image reconstruction and pedestrian re‐identification tasks, independent of the query image resolution. More precisely, an end‐to‐end trainable Resolution Independent person Re‐identification network is proposed. It is composed of our designed Cross‐Resolution GAN and Embedding Batch Normalisation layers. The model is then compared with the traditional low‐resolution pedestrian recognition algorithm and the hybrid method of high‐resolution reconstruction and pedestrian re‐identification. The results demonstrate that the proposed method outperforms the state‐of‐the‐art methods in the pedestrian re‐identification task based on our expanded benchmark dataset. It also reaches an equivalent performance to the existing methods in the high‐resolution image reconstruction task.
Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
