zju3dv/LoFTR issues and pull requests

#142 - 关于Scannet数据集的测试

Issue - State: closed - Opened by Angelo3357 over 2 years ago - 1 comment

#141 - 关于Scannet

Issue - State: closed - Opened by Angelo3357 over 2 years ago - 1 comment

#136 - Provide additional details on HPatches evaluation

Issue - State: open - Opened by Parskatt over 2 years ago - 8 comments

#135 - spv_b_ids issues

Issue - State: closed - Opened by Hoyeon-Ahn over 2 years ago - 1 comment

#134 - Why the training is stopped at the `validation sanity check`

Issue - State: closed - Opened by TruongKhang over 2 years ago - 1 comment

#131 - A possible minor bug in logging

Issue - State: closed - Opened by vdvchen over 2 years ago - 1 comment

#130 - Google colab: google drive link for indoor-ds gives 404

Issue - State: closed - Opened by qwertyforce over 2 years ago - 3 comments

#129 - the number of detect keypoints

Issue - State: closed - Opened by szgy66 over 2 years ago - 3 comments

#128 - I would like to ask whether the path of default.py in the config file needs to be changed after the address is correctly placed. I don't know which ones to modify, but I can't read them now either

Issue - State: closed - Opened by zhubingerdai over 2 years ago - 3 comments

#127 - a question on coarse macth

Issue - State: closed - Opened by ghzhangnj almost 3 years ago - 2 comments

#122 - Hi, how to train the data collected from realsense D435 camera

Issue - State: closed - Opened by ClaudiaShu almost 3 years ago - 1 comment

#117 - A Question about the Image Matching Benchmark 2021 of LoFTR

Issue - State: closed - Opened by Merical almost 3 years ago - 4 comments

#110 - data['spv_b_ids'] 是在哪里定义的,代表什么意思？

Issue - State: closed - Opened by ghzhangnj almost 3 years ago - 1 comment

#105 - how to reproduce the results on hpatches dataset

Issue - State: closed - Opened by Devoe-97 almost 3 years ago - 2 comments

#104 - ScanNet training data

Issue - State: closed - Opened by francescodisalvo05 almost 3 years ago - 1 comment

#103 - About indoor training.

Issue - State: closed - Opened by zwyking almost 3 years ago - 2 comments

#102 - memory consumption exceeds 80gb!!

Issue - State: closed - Opened by sanjaypavo almost 3 years ago - 10 comments

#100 - 关于ScanNet数据集

Issue - State: closed - Opened by Angelo3357 almost 3 years ago - 1 comment

#99 - 关于数据集

Issue - State: closed - Opened by Angelo3357 almost 3 years ago - 1 comment

#98 - a simple question about saving npzfile

Issue - State: closed - Opened by Devoe-97 almost 3 years ago - 16 comments

#96 - Image Stiching

Issue - State: closed - Opened by sivaguru-pat almost 3 years ago - 1 comment

#95 - coarse-matches

Issue - State: closed - Opened by Angelo3357 almost 3 years ago - 1 comment

#94 - 二维特征匹配

Issue - State: closed - Opened by Angelo3357 almost 3 years ago - 1 comment

#93 - dataset preparation for retraining/fine tuning

Issue - State: closed - Opened by whh14 almost 3 years ago - 2 comments

#90 - Training custom datasets with .npz's cause multiple issues with reading the npz files in the dataloader

Issue - State: closed - Opened by amanshenoy almost 3 years ago - 1 comment

#86 - Visualisation of attention weight for linear transformer.

Issue - State: closed - Opened by Miaowshroom almost 3 years ago - 3 comments

#85 - Regarding information required to be able to train on custom dataset

Issue - State: closed - Opened by amanshenoy almost 3 years ago - 9 comments

#84 - Questions about padding for training in coarse matching.

Issue - State: closed - Opened by wangql16 almost 3 years ago - 1 comment

#83 - a simple question about saving npzfile

Issue - State: closed - Opened by Devoe-97 almost 3 years ago - 3 comments

#82 - Megadepth results

Issue - State: closed - Opened by zwyking almost 3 years ago - 12 comments

#81 - Magadepth Weights.

Issue - State: closed - Opened by Qing-WANG-KIT almost 3 years ago - 2 comments

#80 - a question of coarse-matching module

Issue - State: closed - Opened by Angelo3357 almost 3 years ago - 8 comments

#79 - About scannet training

Issue - State: closed - Opened by zwyking almost 3 years ago - 4 comments

#78 - Memory leak.

Issue - State: closed - Opened by Qing-WANG-KIT almost 3 years ago - 3 comments

#74 - Some questions about MegaDepth

Issue - State: closed - Opened by superluckyxhh almost 3 years ago - 2 comments

#71 - There is no dump folder?

Issue - State: closed - Opened by heiyumiao almost 3 years ago - 1 comment

#70 - Aachen数据集测试

Issue - State: closed - Opened by ssssjiang almost 3 years ago - 1 comment

#69 - The memory leakage

Issue - State: closed - Opened by zwyking almost 3 years ago - 2 comments

#68 - Gradio Web Demo

Pull Request - State: closed - Opened by AK391 almost 3 years ago

#67 - Problem Dealing with ScanNet Dataset

Issue - State: closed - Opened by MichiganLu about 3 years ago - 1 comment

#66 - Deployment on KITTI dataset.

Issue - State: closed - Opened by koledickarlo about 3 years ago - 1 comment

#65 - run pretrained model outdoor_ot.ckpt error

Issue - State: closed - Opened by dachengai about 3 years ago - 2 comments

#64 - Update for the trained weights ?

Issue - State: closed - Opened by GabbySuwichaya about 3 years ago - 6 comments

#61 - Aachen test code

Issue - State: closed - Opened by cvlab-flash about 3 years ago - 1 comment

#60 - can you provide data structure for training?

Issue - State: closed - Opened by Xuanyuanbingxue about 3 years ago - 1 comment

#59 - number of matches

Issue - State: closed - Opened by LexaNagiBator228 about 3 years ago - 1 comment

#58 - How to run a batch of image pairs?

Issue - State: closed - Opened by Tourbillon about 3 years ago - 3 comments

#54 - Different image resolution?

Issue - State: closed - Opened by marisancans about 3 years ago - 2 comments

#53 - Some quentions for focal loss

Issue - State: closed - Opened by superluckyxhh about 3 years ago - 3 comments

#52 - Conda pip error and how to fix it

Issue - State: closed - Opened by ducha-aiki about 3 years ago - 1 comment

#49 - Have you used sync BN when training with multiple GPUs?

Issue - State: closed - Opened by Tangshitao about 3 years ago - 1 comment

#48 - Unable to reproduce inference results

Issue - State: closed - Opened by howard5758 about 3 years ago - 12 comments

#47 - Do different colors represent different accuracy rates in your demo?

Issue - State: closed - Opened by Qing-WANG-KIT about 3 years ago - 2 comments

#46 - What is the accuracy if training with 4 GPUs?

Issue - State: closed - Opened by Tangshitao about 3 years ago - 5 comments

#45 - 大神可以把数据放国内一份吗？太难down了，费老劲了，全是泪啊~~

Issue - State: closed - Opened by shanyue-guowei about 3 years ago - 1 comment

#44 - supervision for positions do not have valid depth in GT depth map

Issue - State: closed - Opened by chenghegao4556 about 3 years ago - 1 comment

#43 - Could you explain the training option "focal" ?

Issue - State: closed - Opened by GabbySuwichaya about 3 years ago - 6 comments

#41 - About the positional encoding

Issue - State: closed - Opened by Yang-L1 about 3 years ago - 2 comments

#40 - Auxiliary losses in training statge

Issue - State: closed - Opened by chenghegao4556 about 3 years ago - 4 comments

#39 - Some question about the coarse

Issue - State: closed - Opened by 1962975362 about 3 years ago - 5 comments

#38 - two view image with large parallax

Issue - State: closed - Opened by arenas7307979 about 3 years ago - 1 comment

#37 - Ground-truth labels on my dataset

Issue - State: closed - Opened by 1962975362 about 3 years ago - 6 comments

#36 - Easy way to emulate COTR?

Issue - State: closed - Opened by ducha-aiki about 3 years ago - 2 comments

#35 - Hi, did you train with different window size?

Issue - State: closed - Opened by Devoe-97 about 3 years ago - 1 comment

#34 - Training with single node and multi-GPUs

Issue - State: closed - Opened by drinkingcoder about 3 years ago - 1 comment

#33 - difference between validation results and test results

Issue - State: closed - Opened by zjhthu about 3 years ago - 2 comments

#32 - Adding Mask to Inference

Issue - State: closed - Opened by shlomi-amitai about 3 years ago - 4 comments

#31 - Traning issues

Issue - State: closed - Opened by LexaNagiBator228 about 3 years ago - 1 comment

#30 - the problem during training

Issue - State: closed - Opened by xiezhiqing0112 about 3 years ago - 11 comments

#29 - Problem of ScanNet Dataset pre-processing

Issue - State: closed - Opened by rust-in over 3 years ago - 2 comments

#28 - I want to know the `RESNETFPN.BLOCK_DIMS` while using ResNetFPN with RESOLUTION of (16, 4)

Issue - State: closed - Opened by Devoe-97 over 3 years ago - 2 comments

#27 - Will training code be given these days?

Issue - State: closed - Opened by SSRSYRND over 3 years ago - 3 comments

#26 - Which tool do you use to draw the network pipeline in paper

Issue - State: closed - Opened by HanzhiC over 3 years ago - 1 comment

#25 - Ground truth for training

Issue - State: closed - Opened by GabbySuwichaya over 3 years ago - 4 comments

#22 - Grocery detection

Issue - State: closed - Opened by Grungeby52 over 3 years ago - 1 comment

#21 - Questions on training

Issue - State: closed - Opened by Devoe-97 over 3 years ago - 9 comments

#18 - Questions about names of variables

Issue - State: closed - Opened by park-youngjae over 3 years ago - 2 comments

#17 - Question about Position Encoding

Issue - State: closed - Opened by HanzhiC over 3 years ago - 1 comment

#16 - Could you provide the Scannet training pairs?

Issue - State: closed - Opened by lijiaman over 3 years ago - 3 comments

#15 - Question on bordering operation in CoarseMatching

Issue - State: closed - Opened by codeslake over 3 years ago - 7 comments

#14 - Mutual nearest neighbors

Issue - State: closed - Opened by ChrisLLLLL over 3 years ago - 1 comment

#13 - Training image size

Issue - State: closed - Opened by adizhol over 3 years ago - 1 comment

#12 - Some discussions

Issue - State: closed - Opened by noone-code over 3 years ago - 1 comment

#11 - error "conda env create -f environment.yaml "

Issue - State: closed - Opened by shuttworth over 3 years ago - 1 comment

#10 - Could you provide the training sequence numbers of MegaDepth ?

Issue - State: closed - Opened by noone-code over 3 years ago - 1 comment

#9 - Reproduce Visual Localization Work

Issue - State: closed - Opened by Zhongyuan-Li over 3 years ago - 9 comments

#8 - About the ground-truth labels

Issue - State: closed - Opened by atztao over 3 years ago - 6 comments

#7 - no results appeared !

Issue - State: closed - Opened by trungpham2606 over 3 years ago - 3 comments

#6 - Matching threshold used in training step

Issue - State: closed - Opened by chouscode over 3 years ago - 6 comments

#5 - Poor accuracy

Issue - State: closed - Opened by Zumbalamambo over 3 years ago - 3 comments

#4 - Some quetions about LoFTR

Issue - State: closed - Opened by davidz123 over 3 years ago - 1 comment

#3 - Can you post the code?

Issue - State: closed - Opened by Grungeby52 over 3 years ago - 1 comment

GitHub / zju3dv/LoFTR issues and pull requests