The Importance of Loss Functions for Increasing the Generalization Abilities of a Deep Learning-Based Next Frame Prediction Model for Traffic Scenes
This paper analyzes in detail how different loss functions influence the generalization abilities of a deep learning-based next frame prediction model for traffic scenes. Our prediction model is a convolutional long-short term memory (ConvLSTM) network that generates the pixel values of the next fra...
Main Authors: | Sandra Aigner, Marco Körner |
---|---|
Format: | Article |
Language: | English |
Published: |
MDPI AG
2020-04-01
|
Series: | Machine Learning and Knowledge Extraction |
Subjects: | |
Online Access: | https://www.mdpi.com/2504-4990/2/2/6 |
Similar Items
-
Moving Object Path Prediction in Traffic Scenes Using Contextual Information
by: Jaime B. Fernandez, et al.
Published: (2023-07-01) -
MFVC: Urban Traffic Scene Video Caption Based on Multimodal Fusion
by: Mingxing Li, et al.
Published: (2022-09-01) -
RSG-GCN: Predicting Semantic Relationships in Urban Traffic Scene With Map Geometric Prior
by: Yafu Tian, et al.
Published: (2023-01-01) -
Time-Lag Aware Latent Variable Model for Prediction of Important Scenes Using Baseball Videos and Tweets
by: Kaito Hirasawa, et al.
Published: (2022-03-01) -
Air Traffic Prediction as a Video Prediction Problem Using Convolutional LSTM and Autoencoder
by: Hyewook Kim, et al.
Published: (2021-10-01)