The performance of offline learned pedestrian detectors significantly drops when they are applied to video scenes of various camera views, occlusions, and background structures. Learning a detector for each video scene can avoid the performance drop but it requires repetitive human effort on data annotation. In this paper, a self-learning approach is proposed, toward specifying a pedestrian detector for each video scene without any human annotation involved. Object locations in video frames are treated as latent variables and a progressive latent model (PLM) is proposed to solve such latent variables. The PLM is deployed as components of object discovery, object enforcement, and label propagation, which are used to learn the object locations in a progressive manner. With the difference of convex (DC) objective functions, PLM is optimized by a concave-convex programming algorithm. With specified network branches and loss functions, PLM is integrated with deep feature learning and optimized in an end-to-end manner. From the perspectives of convex regularization and error rate estimation, detailed optimization analysis and learning stability analysis of the proposed PLM are provided. The extensive experiments demonstrate that even without annotation involved the proposed self-learning approach outperforms weakly supervised learning approaches, while achieving comparable performance with transfer learning approaches.


    Zugriff

    Zugriff prüfen

    Verfügbarkeit in meiner Bibliothek prüfen

    Bestellung bei Subito €


    Exportieren, teilen und zitieren



    Titel :

    Progressive Latent Models for Self-Learning Scene-Specific Pedestrian Detectors


    Beteiligte:
    Ye, Qixiang (Autor:in) / Zhang, Tianliang (Autor:in) / Ke, Wei (Autor:in)


    Erscheinungsdatum :

    01.04.2020


    Format / Umfang :

    4280429 byte




    Medientyp :

    Aufsatz (Zeitschrift)


    Format :

    Elektronische Ressource


    Sprache :

    Englisch



    Estimating the Scene-wise Reliability of LiDAR Pedestrian Detectors

    Kyutoku, Haruya / Kawanishi, Yasutomo / Deguchi, Daisuke et al. | IEEE | 2018


    Pedestrian detection by scene dependent classifiers with generative learning

    Yoshida, Hidefumi / Suzuo, Daichi / Deguchi, Daisuke et al. | IEEE | 2013


    PEDESTRIAN DETECTION BY SCENE DEPENDENT CLASSIFIERS WITH GENERATIVE LEARNING

    Yoshida, H. / Suzuo, D. / Deguchi, D. et al. | British Library Conference Proceedings | 2013


    SceneDiffusion: Conditioned Latent Diffusion Models for Traffic Scene Prediction

    Balasubramanian, Lakshman / Wurst, Jonas / Egolf, Robin et al. | IEEE | 2023


    Identifying Latent Classes of Pedestrian Crowd Evacuees

    Haghani, Milad / Sarvi, Majid | Transportation Research Record | 2016