zju3dv/LoFTR issues and pull requests

#318 - tutorial link not working

Issue - State: open - Opened by James-Teng 2 days ago

#317 - Memory Leak?

Issue - State: open - Opened by tomsiu 4 days ago

#316 - D2-Net preprocessed images下载链接失效了

Issue - State: open - Opened by liuqiqi546 about 2 months ago

#315 - utils里面的dataset里面报错这个.client这个代码库没有啊

Issue - State: open - Opened by CruGarnet 2 months ago

#314 - How was pair_infos obtained in the npz file

Issue - State: open - Opened by hbl843449791 3 months ago

#313 - LoFTR: remove unused codes

Pull Request - State: open - Opened by XevenQC 4 months ago

#312 - to reproduce the training,have to download MegaDepth_v1?199G dataset?

Issue - State: open - Opened by pessimstic 4 months ago - 1 comment

#311 - COLMAP dumps for Megadepth 1500 dataset

Issue - State: open - Opened by justachetan 5 months ago

#310 - File "h5py\h5f.pyx", line 96, in h5py.h5f.open OSError: Unable to open file

Issue - State: open - Opened by MintonLee 5 months ago

#309 - ScanNet Training Pairs

Issue - State: open - Opened by justachetan 6 months ago

#308 - My dataset is only RGB, can I use it for training?

Issue - State: open - Opened by liuzihua699 6 months ago

#306 - How to get image feature maps from the pretrained model?

Issue - State: open - Opened by Abbsalehi 6 months ago

#305 - megadepth_indices下载不下来

Issue - State: open - Opened by zyxzyx45 7 months ago - 6 comments

#303 - How to train LoFTR on custom dataset?

Issue - State: open - Opened by jocoskragnus 7 months ago

#302 - main_cfg problem

Issue - State: open - Opened by jocoskragnus 7 months ago

#301 - Bug found in supervision process

Issue - State: open - Opened by HJoonKwon 7 months ago

#300 - image size

Issue - State: open - Opened by LooperzZ 7 months ago

#299 - Michael/initial updates

Pull Request - State: open - Opened by pd-michaelstanley 8 months ago

#298 - meet an error when training loftr: EOFError: Ran out of input

Issue - State: closed - Opened by RunyuZhu 9 months ago - 3 comments

#297 - h5py can't read data

Issue - State: closed - Opened by chen9run 10 months ago

#296 - Can the model be limited to detecting key points in a region of interest?

Issue - State: open - Opened by Abbsalehi 10 months ago

#295 - training on custom dataset

Issue - State: closed - Opened by sunset326 11 months ago - 4 comments

#292 - Dev

Pull Request - State: open - Opened by avichai-airis 11 months ago

#290 - train model for image matching

Issue - State: open - Opened by blackjacket996 about 1 year ago

#289 - How to train the model for HPatches

Issue - State: open - Opened by xmlyqing00 about 1 year ago

#288 - Fix broken kornia tutorial link in README.md

Pull Request - State: open - Opened by AlonOfficial about 1 year ago - 1 comment

#287 - About the dataset indices of Megadepth

Issue - State: open - Opened by AICatYY about 1 year ago - 3 comments

#286 - Update supervision.py

Pull Request - State: closed - Opened by ashutoshsingh0223 about 1 year ago - 1 comment

#285 - About train with fp16

Issue - State: closed - Opened by Master-cai about 1 year ago

#283 - Tensorboard visualization - are the indices correct?

Issue - State: open - Opened by wpacholarz about 1 year ago

#282 - training of smaller model

Issue - State: open - Opened by rahuldeshmukh43 about 1 year ago

#281 - Code about Visual Localization

Issue - State: open - Opened by July0928 about 1 year ago

#279 - Training with corrispondences

Issue - State: open - Opened by fedeloper about 1 year ago

#278 - Training stuck on: "initializing ddp: GLOBAL_RANK"

Issue - State: closed - Opened by nonathecoda over 1 year ago - 4 comments

#277 - About intrinsics of Scannet dataset

Issue - State: open - Opened by sugar-fly over 1 year ago - 1 comment

#276 - How to train in MegaDepth without D2-Net data

Issue - State: open - Opened by QiuhangLiu over 1 year ago - 11 comments

#275 - What is the scale/range of depth used for training on MegaDepth ?

Issue - State: open - Opened by tcourat over 1 year ago

#274 - Link to D2 Net Dataset is broken

Issue - State: closed - Opened by nonathecoda over 1 year ago

#273 - from einops.einops import rearrange ModuleNotFoundError: No module named 'einops'

Issue - State: closed - Opened by monajalal over 1 year ago

#271 - Huggingface Build error Unexpected build error

Issue - State: closed - Opened by yangyi02 over 1 year ago

#270 - What's the difference between outdoor_ds.ckpt and outdoor_ot.ckpt ?

Issue - State: open - Opened by chlg3321 over 1 year ago - 1 comment

#269 - Textureless poor performance compared with SuperGlue

Issue - State: open - Opened by YJonmo over 1 year ago

#268 - About the batch size setting of training

Issue - State: closed - Opened by sugar-fly over 1 year ago

#267 - Killed!!

Issue - State: closed - Opened by chen9run over 1 year ago

#266 - RuntimeError: main thread is not in main loop

Issue - State: closed - Opened by chen9run over 1 year ago - 1 comment

#265 - Data "D2-Net preprocessed images" is unavailable

Issue - State: open - Opened by WallofWonder over 1 year ago - 10 comments

#264 - Lens camera

Issue - State: open - Opened by YJonmo over 1 year ago

#263 - Inference time

Issue - State: open - Opened by YJonmo over 1 year ago - 7 comments

#261 - A confusion about the coarse_matching

Issue - State: open - Opened by BaochunHe over 1 year ago

#260 - Where are the undistorted images provided by the MegaDepth_v1？ MegaDepth_SfM _v1.tar.xz (tar.xz, 667 GB)or MegaDepth_v1.tar.gz(198.80G)？

Issue - State: closed - Opened by chen9run over 1 year ago - 1 comment

#259 - 关于D2Net的未失真图像不再可用的问题

Issue - State: closed - Opened by chen9run over 1 year ago - 2 comments

#257 - about pretrained model

Issue - State: closed - Opened by iron-man45 over 1 year ago

#256 - Use of masks in feature detection

Issue - State: open - Opened by sandeepnmenon over 1 year ago - 3 comments

#255 - c++ inference?

Issue - State: open - Opened by antithing over 1 year ago

#254 - how to create ground truth?

Issue - State: open - Opened by trand2k over 1 year ago - 18 comments

#253 - Reproducing the training results on a megadepth dataset

Issue - State: open - Opened by FlyFish-space over 1 year ago - 23 comments

#252 - Interesting case when matching a small pic to a large pic

Issue - State: open - Opened by xmba15 over 1 year ago

#251 - The link for D2-Net preprocessed images is down

Issue - State: closed - Opened by dadwadw233 over 1 year ago - 1 comment

#250 - Have you ever installed loftr on the PC side, has anyone done such a thing

Issue - State: open - Opened by yeluoo over 1 year ago

#247 - About high-resolution image matching?

Issue - State: open - Opened by chufall over 1 year ago - 1 comment

#246 - 未来会集成进OpenMMLab中吗？

Issue - State: open - Opened by sylarwcy over 1 year ago

#244 - indoor ckpt is not available

Issue - State: open - Opened by SoraCaptain over 1 year ago

#243 - Questions about groundtruth in the training process

Issue - State: closed - Opened by Ov3RCoM3 over 1 year ago - 2 comments

#242 - How to apply the algorithm of loftr to a complete system？

Issue - State: open - Opened by FujiwaraZayako over 1 year ago - 3 comments

#241 - gradient of confidence

Issue - State: open - Opened by weitong8591 over 1 year ago - 4 comments

#240 - Cannot run the matcher in batch mode

Issue - State: open - Opened by naji-s over 1 year ago - 1 comment

#239 - Colab Weights downloading doesnt work

Issue - State: open - Opened by geoeo over 1 year ago - 6 comments

#238 - 尊敬的作者，想请教一下loftr训练是不是必须有深度图信息

Issue - State: closed - Opened by xuechunliang1 over 1 year ago - 6 comments

#237 - Extract feature and key points using the demo

Issue - State: open - Opened by tinaplus almost 2 years ago

#236 - questions on query point refinement.

Issue - State: closed - Opened by minooisbusy almost 2 years ago - 1 comment

#235 - train and test on PhotoTourism dataset

Issue - State: open - Opened by atoshpulatov almost 2 years ago - 12 comments

#234 - Quantifying the LoFTR

Issue - State: open - Opened by HarshitBardana almost 2 years ago - 2 comments

#233 - quantification of the result

Issue - State: open - Opened by ChethanN01 almost 2 years ago

#232 - D2-Net Preprocessed Images drive link broken

Issue - State: closed - Opened by youngyyp almost 2 years ago

#231 - KeyError: 'spv_b_ids'

Issue - State: open - Opened by yash0307 almost 2 years ago

#230 - Megadepth数据集中的深度图解析出来全是0

Issue - State: open - Opened by 18161003959 almost 2 years ago - 2 comments

#229 - D2-Net Preprocessed Images drive link broken

Issue - State: closed - Opened by milesk33 almost 2 years ago - 1 comment

#220 - match threshold

Issue - State: closed - Opened by Lucifer1002 almost 2 years ago - 2 comments

#218 - Running Single Image on LoFTR

Issue - State: open - Opened by atharvjairath about 2 years ago - 1 comment

#214 - Can we direclty use images from Megadepth?

Issue - State: closed - Opened by ewrfcas about 2 years ago - 1 comment

#213 - Robustness against rotation and flipping

Issue - State: closed - Opened by repoit about 2 years ago - 5 comments

#208 - D2-Net preprocessed images 404 error

Issue - State: closed - Opened by sanssouci7 about 2 years ago - 5 comments

#203 - LoFTR descriptors (feature representation retrival)

Issue - State: open - Opened by DavidTu21 about 2 years ago - 5 comments

#197 - Reproducing the result with 4 (or 8) GPUs

Issue - State: open - Opened by sangrockEG about 2 years ago - 2 comments

#190 - 请问用于自己的场景可以直接用预训练模型吗，

Issue - State: closed - Opened by swjtulinxi about 2 years ago - 19 comments

#187 - NameError: name 'conf_min' is not defined

Issue - State: open - Opened by Billy-Mazotti about 2 years ago - 2 comments

#186 - Training on custom datasets

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

#185 - Change the demo query image (更改demo查詢圖像)

Issue - State: open - Opened by lionsjay over 2 years ago - 1 comment

#184 - Different results when changing the input order of image pair

Issue - State: open - Opened by Beniko95J over 2 years ago - 4 comments

GitHub / zju3dv/LoFTR issues and pull requests