Unsupervised Domain Adaptation for Depth Prediction from Images.

08:00 EDT 12th September 2019 | BioPortfolio

Summary of "Unsupervised Domain Adaptation for Depth Prediction from Images."

State-of-the-art methods to infer dense and accurate depth measurements from images rely on deep CNN models trained in an end-to-end fashion on a significant amount of data. However, despite the outstanding performance achieved, these frameworks suffer a drastic drop in accuracy when dealing with unseen environments much different, concerning appearance (e.g., synthetic vs. real) or context (e.g., indoor vs. outdoor), from those observed during the training phase. Such domain shift issue is usually softened by fine-tuning on smaller sets of images with depth labels acquired in the target domain with active sensors (e.g., LiDAR). However, relying on such supervised labeled data is seldom feasible in practical applications. Therefore, we propose an effective unsupervised domain adaptation technique enabling to overcome the domain shift problem without requiring any groundtruth label. Our method, deploying much more accessible to obtain stereo pairs, leverages traditional and not learning-based stereo algorithms to produce disparity/depth labels and on confidence measures to assess their degree of reliability. With these cues, we can fine-tune deep models through a novel confidence-guided loss function, neglecting the effect of outliers gathered from the output of conventional stereo algorithms.


Journal Details

This article was published in the following journal.

Name: IEEE transactions on pattern analysis and machine intelligence
ISSN: 1939-3539


DeepDyve research library

PubMed Articles [10702 Associated PubMed Articles listed on BioPortfolio]

Unsupervised Domain Adaptation for Classification of Histopathology Whole-Slide Images.

Computational image analysis is one means for evaluating digitized histopathology specimens that can increase the reproducibility and reliability with which cancer diagnoses are rendered while simulta...

Inferring Latent Domains for Unsupervised Deep Domain Adaptation.

Unsupervised Domain Adaptation (UDA) refers to the problem of learning a model in a target domain where labeled data are not available by leveraging information from annotated data in a source domain....

Transfer Neural Trees: Semi-Supervised Heterogeneous Domain Adaptation and Beyond.

Heterogeneous domain adaptation (HDA) addresses the task of associating data not only across dissimilar domains but also described by different types of features. Inspired by the recent advances of ne...

Adversarial training with cycle consistency for unsupervised super-resolution in endomicroscopy.

In recent years, endomicroscopy has become increasingly used for diagnostic purposes and interventional guidance. It can provide intraoperative aids for real-time tissue characterization and can help ...

Dual-domain convolutional neural networks for improving structural information in 3 T MRI.

We propose a novel dual-domain convolutional neural network framework to improve structural information of routine 3 T images. We introduce a parameter-efficient butterfly network that involves two ...

Clinical Trials [3319 Associated Clinical Trials listed on BioPortfolio]

Endoscopic Ultrasonographic Staging and Conventional Endoscopic Staging for Depth of Invasion for Early Gastric Cancer

The accurate prediction of depth of tumor invasion in early gastric cancer is essential for the proper selection of candidates for endoscopic resection. Conventional endoscopy and endoscop...

Spatial Frequency Domain Imaging (SFD) for Assessment of Diabetic Foot Ulcer Development and Healing

This study is designed to evaluate the ability of Spatial Frequency Domain Imaging (SFDI) to measure perfusion in lower extremities for prediction of both healing and formation of diabetic...

Multi-Dose Study to Assess the Patients' Ability to Self-Administer Fasinumab in an Unsupervised Setting

The primary objective is to demonstrate that the Prefilled Syringe (PFS) and the auto-injector(AI) are suitable to be used to administer fasinumab at home by patients or their caregivers, ...

Comparison of Supervised and Unsupervised Physical Activity Programs During a Weight Loss Intervention for Adults

Adults who are overweight or obese are typically prescribed a calorie-restricted diet and physical activity to promote weight loss and improve health. The manner in which physical activity...

"New Perspectives of Adaptation to NIV in ALS"

The purpose of this study is to demonstrate that the adaptation to the Non Invasive Ventilation (NIV) at home is not worse when compared with an adaptation performed in inpatient settings.

Medical and Biotech [MESH] Definitions

The adjustment of the eye to variations in the intensity of light. Light adaptation is the adjustment of the eye when the light threshold is increased; DARK ADAPTATION when the light is greatly reduced. (From Cline et al., Dictionary of Visual Science, 4th ed)

The process of generating three-dimensional images by electronic, photographic, or other methods. For example, three-dimensional images can be generated by assembling multiple tomographic images with the aid of a computer, while photographic 3-D images (HOLOGRAPHY) can be made by exposing film to the interference pattern created when two laser light sources shine on an object.

The caring for individuals in institutions and their adaptation to routines characteristic of the institutional environment, and/or their loss of adaptation to life outside the institution.

Predicting the time of OVULATION can be achieved by measuring the preovulatory elevation of ESTRADIOL; LUTEINIZING HORMONE or other hormones in BLOOD or URINE. Accuracy of ovulation prediction depends on the completeness of the hormone profiles, and the ability to determine the preovulatory LH peak.

Combination or superimposition of two images for demonstrating differences between them (e.g., radiograph with contrast vs. one without, radionuclide images using different radionuclides, radiograph vs. radionuclide image) and in the preparation of audiovisual materials (e.g., offsetting identical images, coloring of vessels in angiograms).

Quick Search

DeepDyve research library

Relevant Topic

Antiretroviral therapy
Standard antiretroviral therapy (ART) consists of the combination of at least three antiretroviral (ARV) drugs to maximally suppress the HIV virus and stop the progression of HIV disease. Huge reductions have been seen in rates of death and suffering whe...

Searches Linking to this Article