Analysis of NaN Divergence in Training Monocular Depth Estimation Model
The latest advances in deep learning have facilitated the development of highly accurate monocular depth estimation models. However, when training a monocular depth estimation network, practitioners and researchers have observed not a number (NaN) loss, which disrupts gradient descent optimization. Although several practitioners have reported the stochastic and mysterious occurrence of NaN loss that bothers training, its root cause is not discussed in the literature. This study conducted an in-depth analysis of NaN loss during training a monocular depth estimation network and identified three types of vulnerabilities that cause NaN loss: 1) the use of square root loss, which leads to an unstable gradient; 2) the log-sigmoid function, which exhibits numerical stability issues; and 3) certain variance implementations, which yield incorrect computations. Furthermore, for each vulnerability, the occurrence of NaN loss was demonstrated and practical guidelines to prevent NaN loss were presented. Experiments showed that both optimization stability and performance on monocular depth estimation could be improved by following our guidelines.
Code (0)
등록된 구현이 없습니다.
Tasks
Depth EstimationMonocular Depth EstimationSimilar Papers 제목 키워드 기반
MetricDepth: Enhancing Monocular Depth Estimation with Deep Metric Learning
Deep metric learning aims to learn features relying on the consistency or divergence of class labels. However, in monocular depth estimation, the absence of a natural definition of class poses challenges in the leveragin…
Depth EstimationMetric LearningMonocular Depth EstimationPix2Point: Learning Outdoor 3D Using Sparse Point Clouds and Optimal Transport
Good quality reconstruction and comprehension of a scene rely on 3D estimation methods. The 3D information was usually obtained from images by stereo-photogrammetry, but deep learning has recently provided us with excell…
Deep LearningDepth EstimationMonocular Depth EstimationImage Masking for Robust Self-Supervised Monocular Depth Estimation
Self-supervised monocular depth estimation is a salient task for 3D scene understanding. Learned jointly with monocular ego-motion estimation, several methods have been proposed to predict accurate pixel-wise depth witho…
Autonomous DrivingDepth EstimationMonocular Depth EstimationMotion Estimation+1Monocular Depth Estimation Based On Deep Learning: An Overview
Depth information is important for autonomous systems to perceive environments and estimate their own state. Traditional depth estimation methods, like structure from motion and stereo vision matching, are built on featu…
Deep LearningDepth EstimationMonocular Depth EstimationTwo-in-One Depth: Bridging the Gap Between Monocular and Binocular Self-supervised Depth Estimation
Monocular and binocular self-supervised depth estimations are two important and related tasks in computer vision, which aim to predict scene depths from single images and stereo image pairs respectively. In literature, t…
Depth EstimationMonocular Depth Estimation