Unsupervised Multi-Spectrum Stereo Depth Estimation for All-Day Vision
ieeexplore.ieee.org › document
Nov 9, 2023 · We propose a new unsupervised all-day depth estimation framework with multi-spectrum stereo images, specifically, a thermal image (acting as the right view) ...
Dec 9, 2024 · Simultaneously, we release a large multi-spectrum stereo dataset, which includes TIR-VIS stereo images collected during the day and night.
Feb 23, 2024 · For the multi-spectrum stereo image sequences, we use image sequences of one spe- cific spectrum to estimate the initial depth based on the re-.
In experiments, we demonstrate the better per- formance and generalization of depth estimation through the proposed multispectral stereo dataset, including ...
We propose an efficient and novel multi-task framework called Multispectral Transfer Network (MTN) to estimate depth image from a single thermal image.
Missing: Spectrum | Show results with:Spectrum
Apr 27, 2018 · We propose an efficient and novel multi-task framework called the Multispectral Transfer Network (MTN) to estimate a depth image from a single thermal image.
This work proposes an efficient and novel multi-task framework called the Multispectral Transfer Network (MTN) to estimate a depth image from a single ...
People also ask
What is stereo depth estimation?
What is monocular depth estimation for autonomous driving?
Oct 22, 2024 · Based on an analysis of multispectral properties and the relevance to depth predictions, we propose an efficient and novel multi-task framework ...
Nov 6, 2024 · We propose a novel framework incorporating stereo depth estimation to enforce accurate geometric constraints.
The proposed framework is trained on a rectified multi- spectral stereo dataset which contains 10000 image pairs covering different scenes including campus, ...