Scene detection model training method and device, electronic equipment and storage medium

The embodiment of the invention relates to a scene detection model training method and device, electronic equipment and a storage medium. The method comprises the steps of obtaining a first sample image and annotated scene information; utilizing a pre-trained three-dimensional feature extraction net...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: SU XINGYI, ZHU XINGENG, LI YANG, ZHANG CAO
Format: Patent
Sprache:chi ; eng
Schlagworte:
Online-Zugang:Volltext bestellen
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:The embodiment of the invention relates to a scene detection model training method and device, electronic equipment and a storage medium. The method comprises the steps of obtaining a first sample image and annotated scene information; utilizing a pre-trained three-dimensional feature extraction network contained in the initial scene detection model to perform three-dimensional feature extraction on the at least one first sample image under a target view angle to obtain three-dimensional feature data; performing scene detection on the three-dimensional feature data by using a scene detection network included in the model to obtain predicted scene information; calculating a first loss value representing an error between the predicted scene information and the tagged scene information using a first loss function; adjusting parameters of the scene detection network by using the first loss value; and determining the adjusted initial scene detection model as a trained scene detection model. The model trained by th