Enhanced encoder–decoder architecture for accurate monocular depth estimation

Estimating depth from a single 2D image is a challenging task due to the lack of stereo or multi-view data, which are typically required for depth perception. In state-of-the-art architectures, the main challenge is to efficiently capture complex objects and fine-grained details, which are often dif...

Celý popis

Uloženo v:
Podrobná bibliografie
Vydáno v:The Visual computer Ročník 41; číslo 12; s. 9487 - 9508
Hlavní autoři: Das, Dabbrata, Das, Argho Deb, Sadaf, Farhan
Médium: Journal Article
Jazyk:angličtina
Vydáno: Heidelberg Springer Nature B.V 01.09.2025
Témata:
Accuracy
Algorithms
Augmented reality
Coders
Computer vision
Datasets
Deep learning
Efficiency
Estimation
Neural networks
Real time
Sensors
Space perception
Vision systems
ISSN:0178-2789, 1432-2315
On-line přístup:Získat plný text
Tagy: Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
Popis
Shrnutí:Estimating depth from a single 2D image is a challenging task due to the lack of stereo or multi-view data, which are typically required for depth perception. In state-of-the-art architectures, the main challenge is to efficiently capture complex objects and fine-grained details, which are often difficult to predict. This paper introduces a novel deep-learning-based approach using an enhanced encoder–decoder architecture, where the Inception-ResNet-v2 model serves as the encoder. This is the first instance of utilizing Inception-ResNet-v2 as an encoder for monocular depth estimation, demonstrating improved performance over previous models. It incorporates multi-scale feature extraction to enhance depth prediction accuracy across various object sizes and distances. We propose a composite loss function comprising depth loss, gradient edge loss, and Structural Similarity Index Measure loss, with fine-tuned weights to optimize the weighted sum, ensuring a balance across different aspects of depth estimation. Experimental results on the KITTI dataset show that our model achieves a significantly faster inference time of 0.019 s, outperforming vision transformers in efficiency while maintaining good accuracy. On the NYU-Depth V2 dataset, the model establishes state-of-the-art performance, with an absolute relative error of 0.064, a root-mean-square error of 0.228, and an accuracy of 89.3% for δ < 1.25. These metrics demonstrate that our model can accurately and efficiently predict depth even in challenging scenarios, providing a practical solution for real-time applications.
Bibliografie:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
ISSN:0178-2789
1432-2315
DOI:10.1007/s00371-025-03972-z