Focal-WNet: An Architecture Unifying Convolution and Attention for Depth Estimation

Manimaran, Gouthamaan; Swaminathan, J

doi:10.1109/i2ct54291.2022.9824488

Cited by 7 publications

(4 citation statements)

References 22 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In terms of error performance, our TP-GAN obtains the lowest RMS error by a margin of 0.009 when compared with a very close competitive method of [10]. While our quantitative results are not as good as to those of [51] on the KITTI data, ours outperforms a vast majority of previous methods [7], [11], [12], [16], [45], [46], [47], [48], [49], [50]. Among the approaches, our TP-GAN achieves the best on SQREL and RMSE LOG scores.…”

Section: A Non-adversarial Modelsmentioning

confidence: 64%

“…In terms of accuracy, as demonstrated in Tab. 4, our technique surpasses all nine previous adversarial works [19], [20], [21], [22], [23], [24], [25], [26], [27] as well as non-adversarial methods [7], [11], [12], [16], [45], [46], [47], [48], [49], [50] by significant margins for all the three thresholds δ < 1.25, δ < 1.25 2 , and δ < 1.25 3 , but performs lower than the work in [51] with a small margin. In Tab.…”

Section: ) Kitti Datasetmentioning

confidence: 80%

“…Among the approaches, our TP-GAN achieves the best on SQREL and RMSE LOG scores. In fact, the transformer-based model of [51] employs a far more complex and much deeper network with two distinct encoders and a single decoder. Consequently, the model will have significantly larger parameters and demand much more GPU RAM to train.…”

Section: A Non-adversarial Modelsmentioning

confidence: 99%

See 2 more Smart Citations

TP-GAN: Simple Adversarial Network With Additional Player for Dense Depth Image Estimation

Hendra

Kanazawa

2023

IEEE Access

View full text Add to dashboard Cite

We present a simple yet robust monocular depth estimation technique by synthesizing a depth map image from a single RGB input image using the advantage of generative adversarial networks (GAN). We employ an additional sub-model termed refiner to extract local depth features, then combine it with the global scene information from the generator to improve the GAN's performance compared to the standard GAN architectural scheme. Notably, the generator is the first player to learn to synthesize depth images. The second player, the discriminator, classifies the generated depth. In the meantime, the third player, the refiner, enhances the final reconstructed depth. Complementing the GAN model, we apply a conditional generative network (cGAN) to lead the generator in mapping the input image to the respective depth representation. We further incorporate a structured similarity (SSIM) as our loss function for the generator and refiner in GAN training. Through extensive experiment validation, we confirmed the performance of our strategy on the publicly indoor NYU Depth v2 and KITTI outdoor data. Experiment results on the NYU depth v2 dataset show that our proposed approach achieves the best performance by 96.0% on threshold accuracy (δ < 1.25 2 ) and the second-best accuracy on all thresholds on the KITTI dataset. We discovered that our proposed method compares favorably to numerous existing monocular depth estimation strategies and demonstrates a considerable improvement in the accuracy of image depth estimation despite its simple network architecture.INDEX TERMS Depth estimation, single image, conditional GAN, generative adversarial network (GAN), third player GAN.

show abstract

Section: A Non-adversarial Modelsmentioning

confidence: 64%

Section: ) Kitti Datasetmentioning

confidence: 80%

Section: A Non-adversarial Modelsmentioning

confidence: 99%

See 1 more Smart Citation

TP-GAN: Simple Adversarial Network With Additional Player for Dense Depth Image Estimation

Hendra

Kanazawa

2023

IEEE Access

View full text Add to dashboard Cite

show abstract

“…Then, the features from CNN and Transformer are fed into a cross-attention module to model the dependency between the global and local information. Being similar to that of Depth-Former, Manimaran et al also exploited two separate encoders to extract global and local features [77]. The Transformer-based encoder uses images of 224 × 224 as inputs, while the encoder based on DenseNet [78] uses images of 512 × 512 as inputs.…”

Section: Supervised Trainingmentioning

confidence: 99%

Attention Mechanism Used in Monocular Depth Estimation: An Overview

Li,

Wei,

Fan

2023

Applied Sciences

View full text Add to dashboard Cite

Monocular depth estimation (MDE), as one of the fundamental tasks of computer vision, plays important roles in downstream applications such as virtual reality, 3D reconstruction, and robotic navigation. Convolutional neural networks (CNN)-based methods gained remarkable progress compared with traditional methods using visual cues. However, recent researches reveal that the performance of MDE using CNN could be degraded due to the local receptive field of CNN. To bridge the gap, various attention mechanisms were proposed to model the long-range dependency. Although reviews of MDE algorithms based on CNN were reported, a comprehensive outline of how attention boosts MDE performance is not explored yet. In this paper, we firstly categorize recent attention-related works into CNN-based, Transformer-based, and hybrid (CNN–Transformer-based) approaches in the light of how the attention mechanism impacts the extraction of global features. Secondly, we discuss the details and contributions of attention-based MDE methods published from 2020 to 2022. Then, we compare the performance of the typical attention-based methods. Finally, the challenges and trends of the attention mechanism used in MDE are discussed.

show abstract

Dyna-MSDepth: multi-scale self-supervised monocular depth estimation network for visual SLAM in dynamic scenes

Yao,

Li,

2024

Machine Vision and Applications

View full text Add to dashboard Cite

Focal-WNet: An Architecture Unifying Convolution and Attention for Depth Estimation

Cited by 7 publications

References 22 publications

TP-GAN: Simple Adversarial Network With Additional Player for Dense Depth Image Estimation

TP-GAN: Simple Adversarial Network With Additional Player for Dense Depth Image Estimation

Attention Mechanism Used in Monocular Depth Estimation: An Overview

Dyna-MSDepth: multi-scale self-supervised monocular depth estimation network for visual SLAM in dynamic scenes

Contact Info

Product

Resources

About