Monocular Depth Estimation with Joint Attention Feature Distillation and Wavelet-Based Loss Function

Liu, Peng; Zhang, Zonghua; Meng, Zhaozong; Gao, Nan

doi:10.3390/s21010054

Cited by 13 publications

(6 citation statements)

References 50 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…There are also hybrid methods that use both stereo pair data and video sequence image frames [32,33]. Other refinement strategies, such as [34][35][36]. Some recent approaches used relatively bulky architectures to improve the depth quality [37], which indicates a higher memory cost as well as time.…”

Section: Self-supervised Methodsmentioning

confidence: 99%

The Constraints between Edge Depth and Uncertainty for Monocular Depth Estimation

Liang

et al. 2021

Electronics

View full text Add to dashboard Cite

The self-supervised monocular depth estimation paradigm has become an important branch of computer vision depth-estimation tasks. However, the depth estimation problem arising from object edge depth pulling or occlusion is still unsolved. The grayscale discontinuity of object edges leads to a relatively high depth uncertainty of pixels in these regions. We improve the geometric edge prediction results by taking uncertainty into account in the depth-estimation task. To this end, we explore how uncertainty affects this task and propose a new self-supervised monocular depth estimation technique based on multi-scale uncertainty. In addition, we introduce a teacher–student architecture in models and investigate the impact of different teacher networks on the depth and uncertainty results. We evaluate the performance of our paradigm in detail on the standard KITTI dataset. The experimental results show that the accuracy of our method increased from 87.7% to 88.2%, the AbsRel error rate decreased from 0.115 to 0.11, the SqRel error rate decreased from 0.903 to 0.822, and the RMSE error rate decreased from 4.863 to 4.686 compared with the benchmark Monodepth2. Our approach has a positive impact on the problem of texture replication or inaccurate object boundaries, producing sharper and smoother depth images.

show abstract

Section: Self-supervised Methodsmentioning

confidence: 99%

The Constraints between Edge Depth and Uncertainty for Monocular Depth Estimation

Liang

et al. 2021

Electronics

View full text Add to dashboard Cite

show abstract

“…#params REL↓ RMSE↓ δ1↑ δ2↑ δ3↑ DORN [15] 110.0M 0.072 2.626 0.932 0.984 0.994 AFDB-Net [39] 139.2M 0.071 2.848 0.933 0.983 0.995 VNL [66] 114.2M 0.072 3.258 0.938 0.990 0.998 BTS [34] 112.8M 0.059 2.756 0.956 0.993 0.998 PGA-Net [61] 168.3M 0.063 2.634 0.952 0.992 0.998 TransDepth [65] 311.3M 0.064 2.755 0.956 0.994 0.999 DPT [48] 123…”

Section: Methodsmentioning

confidence: 99%

“…Later on, attention was implemented as spatial attention [1,57], channel-wise attention [22,54], and mix attention [56] to improve classification and detection accuracy. Recent monocular depth estimation methods [25,32,36,39,62] also applied the attention mechanism. However, these attention implementations require relatively heavy computational resources.…”

Section: Related Workmentioning

confidence: 99%

Monocular Depth Estimation Primed by Salient Point Detection and Normalized Hessian Loss

Huynh¹,

Pedone²,

Nguyen³

et al. 2021

Preprint

View full text Add to dashboard Cite

Deep neural networks have recently thrived on single image depth estimation. That being said, current developments on this topic highlight an apparent compromise between accuracy and network size. This work proposes an accurate and lightweight framework for monocular depth estimation based on a self-attention mechanism stemming from salient point detection. Specifically, we utilize a sparse set of keypoints to train a FuSaNet model that consists of two major components: Fusion-Net and Saliency-Net. In addition, we introduce a normalized Hessian loss term invariant to scaling and shear along the depth direction, which is shown to substantially improve the accuracy. The proposed method achieves state-of-the-art results on NYU-Depth-v2 and KITTI while using 3.1-38.4 times smaller model in terms of the number of parameters than baseline approaches. Experiments on the SUN-RGBD further demonstrate the generalizability of the proposed method.

show abstract

“…Recently, Refs. [ 31 , 41 , 46 ] studied wavelet–based loss function to improve structural details while [ 32 ] proposed learning sparse wavelet coefficients to estimate depth maps. Researchers have also replaced the UNET encoder backbone with dense pretrained image networks, such as ResNet, DenseNet [ 7 ], etc., and tuned the decoder accordingly to estimate depth.…”

Section: Nested Dwt Net Architecturementioning

confidence: 99%

Nested DWT–Based CNN Architecture for Monocular Depth Estimation

Paul

Mishra

Marimuthu

2023

Sensors

View full text Add to dashboard Cite

Applications such as medical diagnosis, navigation, robotics, etc., require 3D images. Recently, deep learning networks have been extensively applied to estimate depth. Depth prediction from 2D images poses a problem that is both ill–posed and non–linear. Such networks are computationally and time–wise expensive as they have dense configurations. Further, the network performance depends on the trained model configuration, the loss functions used, and the dataset applied for training. We propose a moderately dense encoder–decoder network based on discrete wavelet decomposition and trainable coefficients (LL, LH, HL, HH). Our Nested Wavelet–Net (NDWTN) preserves the high–frequency information that is otherwise lost during the downsampling process in the encoder. Furthermore, we study the effect of activation functions, batch normalization, convolution layers, skip, etc., in our models. The network is trained with NYU datasets. Our network trains faster with good results.

show abstract

Monocular Depth Estimation with Joint Attention Feature Distillation and Wavelet-Based Loss Function

Cited by 13 publications

References 50 publications

The Constraints between Edge Depth and Uncertainty for Monocular Depth Estimation

The Constraints between Edge Depth and Uncertainty for Monocular Depth Estimation

Monocular Depth Estimation Primed by Salient Point Detection and Normalized Hessian Loss

Nested DWT–Based CNN Architecture for Monocular Depth Estimation

Contact Info

Product

Resources

About