资源分类
》 资源 》算法》 depth_estimation_with_densenet-unet_hybrid

depth_estimation_with_densenet-unet_hybrid

2020-03-30 | |  37 |   0 |   0

Hybrid DenseNet-based CNN for Single Image Depth Estimation

This repository contains an updated version of the CNN from my repository DE_resnet_unet_hyb. The ResNet backbone has been replaced with a DenseNet169.

Requirements

The code was tested with:

  • python 3.5 and 3.6

  • pytorch (and torchvision) 1.3.0

  • opencv-python 3.4.3

  • matplotlib 2.2.3

  • numpy 1.15.4

Guide

  • Predicting the depth of an arbitrary image:

python3 predict_img.py -i <path_to_image>

Evalutation

  • Quantitative results on the NYU depth v2 test set:

RELRMSELog10δ1δ2δ3
0.1290.5880.0560.8330.9620.990


上一篇：DenseNet_centerloss-CapsulNet-InceptionV3_use

下一篇：ResNet-DenseNet-Notebook

用户评价
全部评价

热门资源

  • seetafaceJNI

    项目介绍 基于中科院seetaface2进行封装的JAVA...

  • spark-corenlp

    This package wraps Stanford CoreNLP annotators ...

  • Keras-ResNeXt

    Keras ResNeXt Implementation of ResNeXt models...

  • capsnet-with-caps...

    CapsNet with capsule-wise convolution Project ...

  • shih-styletransfer

    shih-styletransfer Code from Style Transfer ...