Official repo for the work titled "SharinGAN: Combining Synthetic and Real Data for Unsupervised GeometryEstimation"

Koutilya PNVR

Last update: Oct 19, 2022

Related tags

Deep Learning SharinGAN

Overview

SharinGAN

Official repo for the work titled "SharinGAN: Combining Synthetic and Real Data for Unsupervised GeometryEstimation"

The official project website for this work can be found here.

Requirements

Python 2.7 Pytorch 0.4.1

The trained model files for both the tasks are made available here here. The pretrained_models folder contains the pretrained model for the generator and the primary task networks before end-to-end training the SharinGAN network as a whole. The final trained models are present in Face_Normal_Estimation/ and Monocular_Depth_Estimation/ directories of the google drive.

The environment.yml file is also provided for one to replicate the environment.

We added the training and validation codes for both the tasks of Monocular Depth Estimation and Face Normal Estimation. We hope to improve the repository with time. We appreciate your inputs and feedback

Monocular Depth Estimation

Place the saved model file (Depth_Estimator_WI_geom_bicubic_da-144999.pth.tar) inside a newly created folder Monocular_Depth_Estimation/saved_models/ of the current repo.

The dataset files required for the dataloaders Kitti_dataloader.py and VKitti_dataloader.py are made available at Monocular_Depth_Estimation/dataset_files/.

Place the Monocular_Depth_Estimation/dataset_files/Kitti/.txt files in the original downloaded kitti/ dataset folder. Similarly place the Monocular_Depth_Estimation/dataset_files/VKitti/.txt files in the original downloaded Virtual_Kitti/ dataset folder.

Make3D evaluation

cd Monocular_Depth_Estimation
python Make3D_validation.py --iter 144999

Comments

About Kitti data

Hi, I'm struggling with the Kitti data u used in ur project, can u plz tell me which item did u download from the official Kitti website? Maybe the data u used is Kitti 360? But I didn't find data which is labeled as 2011. Thank you so much for ur help.

opened by yuehua-Song666 5
More details are expected

Thanks for your contribution to this topic. Could you provide more details over your code. For example, the version of Pytorch, or a basic sample run using default parameters.

opened by sudalvxin 3
Comparing Different models

Hi, Could you please tell me how you compared different models? Did you use the same learning rate, number of epochs, Number of decay epochs, image size, optimizer among all models? Also, did you collect test results using the final saved generator or did you use the best results testing all saved generators at different epochs?

opened by mohammadshahabuddin 1
Looking for trained SharinGAN depth estimation model [KITTI]

Hi Koutilya,

Thanks for your contribution to this topic.

With the current script published in Git repo, we notice several bugs, and majorly for reproducing the paper results the important details are missing. [as mentioned in #2 also]

Can you please update us when the trained Sharingan models and/or the bug-free scripts will be available?

Thank you,

opened by Akhil-Gurram 1
about the Kitti and vkitti dataset

Thank you very much for your excellent job and I think it is interesting. I have some trouble about the datasets, although I have download the Kitti and vkitty. Can you give me some help about how to prepare the datasets ?

opened by hello-trouble 1
the trained model for Face_Normal_Estimation

Hi,

Great work~

Could you please share the trained model of Face_Normal_Estimation? I can only find the trained model for Monocular_Depth_Estimation in the Google Driver: https://drive.google.com/drive/folders/1SRznz7AlezF655doEZSAxk_YFSdGGhd4，but nothing for Face_Normal_Estimation.

Thank you very much~

opened by reallm 0
Questions on the training time

Thanks for the great work!

May I ask what is the estimated training time for each stage, i.e., pretraining G, pretraining T, and jointly training G and T, on a single GPU like TITAN or V100? Thanks! : )

opened by SenZHANG-GitHub 0

Official repo for the work titled "SharinGAN: Combining Synthetic and Real Data for Unsupervised GeometryEstimation"

Related tags

Overview

SharinGAN

Requirements

Monocular Depth Estimation

Make3D evaluation

Comments

About Kitti data

More details are expected

Comparing Different models

Looking for trained SharinGAN depth estimation model [KITTI]

about the Kitti and vkitti dataset

the trained model for Face_Normal_Estimation

Questions on the training time

Owner

Koutilya PNVR

Official Repo of my work for SREC Nandyal Machine Learning Bootcamp

We have implemented shaDow-GNN as a general and powerful pipeline for graph representation learning. For more details, please find our paper titled Deep Graph Neural Networks with Shallow Subgraph Samplers, available on arXiv (https//arxiv.org/abs/2012.01380).

Code for reproducing our analysis in the paper titled: Image Cropping on Twitter: Fairness Metrics, their Limitations, and the Importance of Representation, Design, and Agency

This repository contains the code, data, and models of the paper titled "XL-Sum: Large-Scale Multilingual Abstractive Summarization for 44 Languages" published in Findings of the Association for Computational Linguistics: ACL 2021.

Pre-trained BERT Models for Ancient and Medieval Greek, and associated code for LaTeCH 2021 paper titled - "A Pilot Study for BERT Language Modelling and Morphological Analysis for Ancient and Medieval Greek"

The implementation of our CIKM 2021 paper titled as: "Cross-Market Product Recommendation"

This is the repo for our work "Towards Persona-Based Empathetic Conversational Models" (EMNLP 2020)

Official code of our work, Unified Pre-training for Program Understanding and Generation [NAACL 2021].

Official code of our work, AVATAR: A Parallel Corpus for Java-Python Program Translation.

Official Repo for Ground-aware Monocular 3D Object Detection for Autonomous Driving

This repo provides the official code for TransBTS: Multimodal Brain Tumor Segmentation Using Transformer (https://arxiv.org/pdf/2103.04430.pdf).

Official repo for QHack—the quantum machine learning hackathon

The official repo of the CVPR2021 oral paper: Representative Batch Normalization with Feature Calibration

This Repo is the official CUDA implementation of ICCV 2019 Oral paper for CARAFE: Content-Aware ReAssembly of FEatures

Official repo for AutoInt: Automatic Integration for Fast Neural Volume Rendering in CVPR 2021

This repo contains the official code and pre-trained models for the Dynamic Vision Transformer (DVT).

This repo contains the official implementations of EigenDamage: Structured Pruning in the Kronecker-Factored Eigenbasis

Official repo for SemanticGAN https://nv-tlabs.github.io/semanticGAN/