Papers › SwinMTL: A Shared Architecture for Simultaneous Depth Estimation and Semantic...

SwinMTL: A Shared Architecture for Simultaneous Depth Estimation and Semantic Segmentation from Monocular Camera Images

15 Mar 2024arXiv:2403.10662archive 2025-07-28

Pardis Taghavi, Reza Langari, Gaurav Pandey

This research paper presents an innovative multi-task learning framework that allows concurrent depth estimation and semantic segmentation using a single camera. The proposed approach is based on a shared encoder-decoder architecture, which integrates various techniques to improve the accuracy of the depth estimation and semantic segmentation task without compromising computational efficiency. Additionally, the paper incorporates an adversarial training component, employing a Wasserstein GAN framework with a critic network, to refine model's predictions. The framework is thoroughly evaluated on two datasets - the outdoor Cityscapes dataset and the indoor NYU Depth V2 dataset - and it outperforms existing state-of-the-art methods in both segmentation and depth estimation tasks. We also conducted ablation studies to analyze the contributions of different components, including pre-training strategies, the inclusion of critics, the use of logarithmic depth scaling, and advanced image augmentations, to provide a better understanding of the proposed framework. The accompanying source code is accessible at \url{https://github.com/PardisTaghavi/SwinMTL}.

PaperPDFCode

In Syntology Open this paper in Syntology's Atlas, the map of the papers in Syntology's graph and their citations.

Code

pardistaghavi/swinmtl officialmentioned in papermentioned on GitHubpytorch report

Repository list and official/mentioned flags are the archive's, frozen 2025-07-28. Reachability, where shown, is from one Syntology probe window (2026-09-16 to 2026-09-18); repositories not probed show nothing. GitHub stars are not tracked.

Code Syntology ran Syntology

Not run by Syntology. Nothing on this page verifies that the listed code works.

Tasks

Computational EfficiencyDecoderDepth EstimationMonocular Depth EstimationMulti-Task LearningReal-Time Semantic SegmentationSegmentationSemantic Segmentation

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Depth Estimation Cityscapes test SwinMTL RMSE 6.352 #1 of 2 Archive leaderboard report
Monocular Depth Estimation Cityscapes SwinMTL Absolute relative error (AbsRel) 0.089 #1 of 3 Archive leaderboard report
Monocular Depth Estimation Cityscapes SwinMTL RMSE 5.481 #1 of 3 Archive leaderboard report
Monocular Depth Estimation Cityscapes SwinMTL RMSE log 0.139 #1 of 3 Archive leaderboard report
Monocular Depth Estimation Cityscapes SwinMTL Square relative error (SqRel) 1.051 #1 of 3 Archive leaderboard report
Multi-Task Learning Cityscapes test SwinMTL RMSE 0.51 #1 of 3 Archive leaderboard report
Multi-Task Learning Cityscapes test SwinMTL mIoU 76.41 #1 of 3 Archive leaderboard report
Multi-Task Learning NYUv2 SwinMTL Mean IoU 58.14 #1 of 2 Archive leaderboard report
Real-Time Semantic Segmentation Cityscapes test SwinMTL mIoU 76.41% #9 of 39 Archive leaderboard report
Semantic Segmentation Cityscapes test SwinMTL Mean IoU (class) 76.41% #65 of 105 Archive leaderboard report
Semantic Segmentation Cityscapes val SwinMTL mIoU 76.41 #68 of 99 Archive leaderboard report
Semantic Segmentation NYU Depth v2 SwinMTL Mean IoU 58.14% #11 of 121 Archive leaderboard report

Ranks are positions in the archive's leaderboards as they stood at the 2025-07-28 snapshot. Results published since then are not among these rows, so a rank here is not a current standing.

Methods

Absolute Position EncodingsAdamAttentionBPEDense ConnectionsDropoutLabel SmoothingLayer NormalizationLinear LayerMulti-Head AttentionPosition-Wise Feed-Forward LayerResidual ConnectionSoftmaxTransformer

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections