Ecosyste.ms: Issues

An open API service for providing issue and pull request metadata for open source projects.

GitHub / cvg/glue-factory issues and pull requests

#113 - Training on Custom Dataset

Issue - State: open - Opened by Shafaq-S 11 days ago

#103 - When are you releasing the baseline for RoMa?

Issue - State: open - Opened by nam1410 3 months ago

#102 - LoFTR Evaluation

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

#101 - Can a pre-trained model be called?

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

#100 - fix: numpy api to torch api

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

#98 - Pold2 line fix

Pull Request - State: closed - Opened by RR-2000 5 months ago

#97 - augmenting own data

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

#96 - Pold2 Merge

Pull Request - State: closed - Opened by RR-2000 5 months ago

#95 - Downloading HPatches hangs

Issue - State: open - Opened by axbycc-mark 5 months ago

#94 - End-to-end SP+LG

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

#93 - OmniGlue

Issue - State: open - Opened by Eunchan24 5 months ago - 1 comment

#92 - Slow Training Speed Issue

Issue - State: open - Opened by langy888 5 months ago - 1 comment

#91 - Train

Issue - State: open - Opened by LXXIANG12 5 months ago

#90 - Instability of LightGlue training

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

#88 - The Required Datasets for Training.

Issue - State: open - Opened by zhenpengju 6 months ago - 1 comment

#87 - Performance

Issue - State: open - Opened by junwei0812 7 months ago

#86 - the two places name the gamma differently?

Issue - State: closed - Opened by Devoe-97 7 months ago

#85 - Fix DISK extraction.

Pull Request - State: open - Opened by hit2sjtu 7 months ago

#84 - Merge gt generation

Pull Request - State: closed - Opened by ewgoedeke 7 months ago

#83 - simpler GT line matching

Issue - State: closed - Opened by Golbstein 7 months ago - 1 comment

#82 - Eg gt testing

Pull Request - State: closed - Opened by ewgoedeke 7 months ago

#81 - gluestick training time

Issue - State: closed - Opened by fatih4d 7 months ago - 3 comments

#79 - Draft: Rk 1 network and pipeline

Pull Request - State: closed - Opened by RaphaelKreft 7 months ago - 1 comment

#78 - what is add_scale_ori and how to use it?

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

#76 - mixed precision training

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

#75 - memory not release

Issue - State: closed - Opened by junwei0812 8 months ago - 3 comments

#74 - DeDoDe

Issue - State: open - Opened by terrance-wang 8 months ago - 1 comment

#73 - pycolmap error

Issue - State: closed - Opened by zj12345648 8 months ago - 1 comment

#72 - train => disk +lg

Issue - State: open - Opened by vampire1204 8 months ago - 1 comment

#71 - How to train lightglue with my extractor?

Issue - State: open - Opened by Adolfhill 8 months ago - 2 comments

#69 - what is the approximate training time for each stages?

Issue - State: open - Opened by Moreland-cas 8 months ago - 1 comment

#68 - Allow SuperPoint_Open weights to be loaded from local cache

Pull Request - State: closed - Opened by aduverger 8 months ago - 2 comments

#66 - question about homography transformations

Issue - State: closed - Opened by noahzn 9 months ago - 2 comments

#65 - Why duplicating the log_assignment calculation?

Issue - State: open - Opened by 013292 10 months ago - 1 comment

#64 - Mismatched model shape upon MegaDepth finetuning

Issue - State: closed - Opened by mattiasmar 10 months ago - 2 comments

#63 - Questions about SIFT+LightGlue

Issue - State: closed - Opened by Uchan1996 10 months ago - 1 comment

#62 - KeyError: 'loss/total'

Issue - State: closed - Opened by QuangVinh1280 10 months ago - 7 comments

#61 - Pre-training Error

Issue - State: closed - Opened by QuangVinh1280 10 months ago

#60 - Struture of the MegaDepth dataset

Issue - State: open - Opened by QuangVinh1280 10 months ago - 1 comment

#59 - Add option to chain multiple learning rate schedulers.

Pull Request - State: open - Opened by veichta 10 months ago

#58 - while Reproducing... can't reach the same performance

Issue - State: open - Opened by AubreyCH 10 months ago - 3 comments

#57 - error when installing

Issue - State: open - Opened by 2659494286 10 months ago

#56 - Added DoG-HardNet extractor and configs

Pull Request - State: open - Opened by ducha-aiki 10 months ago

#55 - Drop last in distributed mode

Pull Request - State: closed - Opened by sarlinpe 11 months ago

#54 - Trained Weights

Issue - State: open - Opened by flashcolab 11 months ago - 2 comments

#53 - Homography pretraining not converging

Issue - State: closed - Opened by ikaftan 11 months ago

#52 - Adds evaluation on challenging WxBS and EVD datasets

Pull Request - State: open - Opened by ducha-aiki 11 months ago - 1 comment

#51 - Added KeyNetAffNetHardNet model

Pull Request - State: closed - Opened by ducha-aiki 11 months ago - 1 comment

#50 - Fix SIFT training

Pull Request - State: closed - Opened by hit2sjtu 11 months ago

#49 - A failed attempt to use official PTH as the initial training value

Issue - State: open - Opened by Zhaoyibinn 11 months ago - 6 comments

#48 - Conversion of trained checkpoint to official lg checkpoint

Issue - State: closed - Opened by spagnoloG 11 months ago - 6 comments

#47 - KeyError: 'loss/total'

Issue - State: closed - Opened by flashcolab 12 months ago - 4 comments

#46 - Fix minor typo in NLL

Pull Request - State: closed - Opened by sarlinpe 12 months ago

#44 - Possible typo?

Issue - State: closed - Opened by AlbertoJaenal 12 months ago - 1 comment

#43 - Discrepancy in training config w.r.t. paper

Issue - State: open - Opened by ducha-aiki 12 months ago

#42 - question about the relative_pose_error

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

#41 - I have a problem after training is complete

Issue - State: open - Opened by qiao89757 about 1 year ago - 1 comment

#40 - Training SuperGlue on custom data

Issue - State: open - Opened by Shuhul24 about 1 year ago - 6 comments

#39 - Possibly faster training

Issue - State: open - Opened by ducha-aiki about 1 year ago - 4 comments

#38 - Fixing #37 allow_shifted=False in MotionBlur

Pull Request - State: closed - Opened by iago-suarez about 1 year ago

#37 - Danger: Albumentations MotionBlur is shifting our images

Issue - State: closed - Opened by iago-suarez about 1 year ago
Labels: bug

#36 - About training LightGlue in two stages

Issue - State: open - Opened by EATMustard about 1 year ago - 1 comment

#35 - Eval on WxBS

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

#34 - Training on my own dataset

Issue - State: open - Opened by CalleRC about 1 year ago - 6 comments

#33 - Questions about training data

Issue - State: closed - Opened by ACuOoOoO about 1 year ago - 1 comment

#32 - Fix typo in train.py

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

#31 - Issues with lrscheduler

Issue - State: closed - Opened by AnasAlsharo about 1 year ago - 1 comment

#30 - Avoid unnecessary generation and storage of figures during training

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

#29 - Fix GlueStick training config

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

#28 - About train

Issue - State: open - Opened by liutao23 about 1 year ago - 3 comments

#27 - Fix config links in README

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

#26 - SIFT+ALIKED results

Pull Request - State: closed - Opened by Phil26AT about 1 year ago

#25 - Add Scannet1500 dataset

Pull Request - State: open - Opened by Phil26AT about 1 year ago

#24 - Feature/pytheia

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

#23 - Performance differences between 2 versions of SuperPoint with pretrained weights

Issue - State: closed - Opened by nobobobo about 1 year ago - 5 comments

#22 - LightGlue weights trained on SuperPoint-Open features

Issue - State: open - Opened by fabio-sim about 1 year ago - 2 comments

#21 - Any numbers for new models (open SP, ALIKED, SIFT)?

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

#18 - Add training configs for SuperPoint (MagicLeap) and DISK with LightGlue

Pull Request - State: closed - Opened by Phil26AT about 1 year ago

#17 - Adding basic CI and fixing minor problems

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

#16 - About installation

Issue - State: closed - Opened by PeeKaBo0L about 1 year ago - 4 comments

#14 - Homography disk requirement

Issue - State: closed - Opened by sbucaille about 1 year ago - 2 comments

#13 - Avoid sparse depth export for megadepth

Pull Request - State: closed - Opened by Phil26AT about 1 year ago

#11 - Add flexible load for lr scheduler.

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

#9 - Check for model initialization during eval

Pull Request - State: closed - Opened by Phil26AT about 1 year ago

#8 - Create data folder before downloading MegaDepth1500

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

#7 - errors when evaling on datasets

Issue - State: closed - Opened by kajo-kurisu about 1 year ago - 4 comments

#6 - Auto sort imports

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

#5 - Cleanup install and dependencies

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

#4 - No module named 'resource'

Issue - State: closed - Opened by UyCode about 1 year ago - 3 comments

#3 - Add dev dependencies

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

#2 - Bugfix when downloading the homography dataset

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