Abstract: Scene decomposition into its illuminant, shading, and reflectance intrinsic images is an essential step for scene understanding. Collecting intrinsic image groundtruth data is a laborious task. The assumptions on which the ground-truth
procedures are based limit their application to simple scenes with a single object taken in the absence of indirect lighting and interreflections. We investigate synthetic data for intrinsic image research since the extraction of ground truth is straightforward, and it allows for scenes in more realistic situations (e.g, multiple illuminants and interreflections). With this dataset we aim to motivate researchers to further explore intrinsic image decomposition in complex scenes.

Abstract: We call on the Document AI (DocAI) community to re-evaluate current methodologies and embrace the challenge of creating more practically-oriented benchmarks. Document Understanding Dataset and Evaluation (DUDE) seeks to remediate the halted research progress in understanding visually-rich documents (VRDs). We present a new dataset with novelties related to types of questions, answers, and document layouts based on multi-industry, multi-domain, and multi-page VRDs of various origins and dates. Moreover, we are pushing the boundaries of current methods by creating multi-task and multi-domain evaluation setups that more accurately simulate real-world situations where powerful generalization and adaptation under low-resource settings are desired. DUDE aims to set a new standard as a more practical, long-standing benchmark for the community, and we hope that it will lead to future extensions and contributions that address real-world challenges. Finally, our work illustrates the importance of finding more efficient ways to model language, images, and layout in DocAI.

Abstract: In civilian video security monitoring, retrieving and tracking a person of interest often rely on witness testimony and their appearance description. Deployed systems rely on a large amount of annotated training data and are expected to show consistent performance in diverse areas and gen-eralize well between diverse settings w.r.t. different view-points, illumination, resolution, occlusions, and poses for indoor and outdoor scenes. However, for such generalization, the system would require a large amount of various an-notated data for training and evaluation. The WACV 2023 Pedestrian Attribute Recognition and Attributed-based Per-son Retrieval Challenge (UPAR-Challenge) aimed to spot-light the problem of domain gaps in a real-world surveil-lance context and highlight the challenges and limitations of existing methods. The UPAR dataset, composed of 40 important binary attributes over 12 attribute categories across four datasets, was extended with data captured from a low-flying UAV from the P-DESTRE dataset. To this aim, 0.6M additional annotations were manually labeled and vali-dated. Each track evaluated the robustness of the competing methods to domain shifts by training on limited data from a specific domain and evaluating using data from unseen do-mains. The challenge attracted 41 registered participants, but only one team managed to outperform the baseline on one track, emphasizing the task's difficulty. This work de-scribes the challenge design, the adopted dataset, obtained results, as well as future directions on the topic.

Abstract: Following previous series on Looking at People (LAP) challenges [6, 5, 4], ChaLearn ran two competitions to be presented at CVPR 2015: action/interaction spotting and cultural event recognition in RGB data. We ran a second round on human activity recognition on RGB data sequences. In terms of cultural event recognition, tens of categories have to be recognized. This involves scene understanding and human analysis. This paper summarizes the two performed challenges and obtained results. Details of the ChaLearn LAP competitions can be found at http://gesture.chalearn.org/.

Abstract: Pain is a vital sign of human health and its automatic detection can be of crucial importance in many different contexts, including medical scenarios. While most available computer vision techniques are based on RGB, in this paper, we investigate the effect of combining RGB, depth, and thermal
facial images for pain detection and pain intensity level recognition. For this purpose, we extract energies released by facial pixels using a spatiotemporal filter. Experiments on a group of 12 elderly people applying the multimodal approach show that the proposed method successfully detects pain and recognizes between three intensity levels in 82% of the analyzed frames improving more than 6% over RGB only analysis in similar conditions.

Abstract: In this paper we present a novel polyp region segmentation method for colonoscopy videos. Our method uses valley information associated to polyp boundaries in order to provide an initial segmentation. This first segmentation is refined to eliminate boundary discontinuities caused by image artifacts or other elements of the scene. Experimental results over a publicly annotated database show that our method outperforms both general and specific segmentation methods by providing more accurate regions rich in polyp content. We also prove how image preprocessing is needed to improve final polyp region segmentation.