princeton-vl/RAFT issues and pull requests

#100 - Trian model with masked custom dataset

Issue - State: open - Opened by Roypic over 3 years ago - 4 comments

#100 - Trian model with masked custom dataset

Issue - State: open - Opened by Roypic over 3 years ago - 4 comments

#99 - Dataset order for training schedule

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

#98 - When evaluate on the sintel train datasets,the epe result is nan

Issue - State: closed - Opened by ljp1997 over 3 years ago - 1 comment

#97 - Gray scale image.

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

#97 - Gray scale image.

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

#96 - Different size between optical flow and original image when running demo.py

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

#95 - batch normalization in the update block

Issue - State: closed - Opened by jsczzzk over 3 years ago

#94 - Regarding align_corners=True

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

#93 - RAFT small

Issue - State: open - Opened by antoniopavlis over 3 years ago - 4 comments

#92 - How do I use this alt_cuda_corr during train

Issue - State: open - Opened by aiforworlds over 3 years ago - 2 comments

#91 - How the correlation lookup differs from the correlation layer in Flownet?

Issue - State: closed - Opened by jingjin25 over 3 years ago - 2 comments

#90 - FlyingThings3D Reproduction

Issue - State: open - Opened by gxd1994 over 3 years ago - 1 comment

#89 - question about index correlation volume

Issue - State: closed - Opened by jucic over 3 years ago

#88 - AlternateCorrBlock: fmap1_i definition has a bug?

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

#87 - Motion Blur Real-Data and Baseline

Issue - State: closed - Opened by JonasFrey96 over 3 years ago - 2 comments

#86 - Running the demo with --alternate_corr fails

Issue - State: open - Opened by christian-rauch over 3 years ago - 8 comments

#85 - python3 shebang and whitespace cleanup

Pull Request - State: open - Opened by christian-rauch over 3 years ago - 1 comment

#84 - Training a model in Self-supervised manner

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

#83 - RuntimeError: grid_sampler()

Issue - State: closed - Opened by vanderbilt2 over 3 years ago - 2 comments

#82 - Question about the pretrained sintel model

Issue - State: closed - Opened by Puppet097 over 3 years ago - 1 comment

#81 - Running on 1 GPU

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

#80 - Bilinear sampler bug

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

#79 - Occlusions

Issue - State: closed - Opened by zisismp4 over 3 years ago - 3 comments

#78 - kitti_15 and kitti_12 datasets

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

#77 - RAFT-small

Issue - State: closed - Opened by gxd1994 over 3 years ago

#76 - Training RAFT-small takes longer than RAFT

Issue - State: open - Opened by enric1994 over 3 years ago

#75 - Minimum input image size?

Issue - State: closed - Opened by Sino0904 over 3 years ago - 4 comments

#74 - RAFT-3D

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

#73 - Question about the update operator

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

#72 - some question about Performance measurement

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

#71 - cannot load model

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

#70 - Saw-tooth Effect on high-res images

Issue - State: open - Opened by YangLiu14 over 3 years ago - 1 comment

#69 - Update conda install instructions

Pull Request - State: closed - Opened by JonathonLuiten over 3 years ago

#68 - Question about the training dataset

Issue - State: open - Opened by Puppet097 over 3 years ago

#67 - which dataset you use in your pretrained model??

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

#66 - : cannot connect to X server

Issue - State: open - Opened by aertist over 3 years ago

#65 - Features for Refinement

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

#64 - Backward Warp (use flow_t->t+1 and frame_t+1 to recover frame_t to ensure no undefined area)

Issue - State: open - Opened by lxa9867 almost 4 years ago - 2 comments

#63 - predicting next frame

Issue - State: open - Opened by zisismp4 almost 4 years ago - 4 comments

#62 - validation set and warm start

Issue - State: closed - Opened by az-ja almost 4 years ago - 2 comments

#61 - training with train_mixed

Issue - State: open - Opened by az-ja almost 4 years ago - 1 comment

#60 - In cuda11.0, running time become slower than cuda10.x

Issue - State: open - Opened by JunHeum almost 4 years ago - 3 comments

#59 - About Correlation Range

Issue - State: closed - Opened by BB88Lee almost 4 years ago - 1 comment

#58 - Error while running the demo - Possible OpenCV or PyQt issue

Issue - State: closed - Opened by BarathwajAnandan almost 4 years ago - 1 comment

#57 - flow on occluded areas

Issue - State: closed - Opened by maorp almost 4 years ago - 11 comments

#56 - Different size between optical flow and original image

Issue - State: closed - Opened by 574411705 almost 4 years ago - 1 comment

#55 - Can't download pretrain model

Issue - State: closed - Opened by mosjoker almost 4 years ago - 1 comment

#54 - Why not use amp in CorrBlock?

Issue - State: closed - Opened by chenerg almost 4 years ago - 1 comment

#53 - Colab demo?

Issue - State: open - Opened by noobtoob4lyfe almost 4 years ago - 1 comment

#52 - What is the output of bilinear sampling in each multiscale level?

Issue - State: open - Opened by Starboy-at-earth almost 4 years ago

#51 - demo low resolution optical flow

Issue - State: open - Opened by Pavelrst almost 4 years ago - 1 comment

#50 - demo does not work

Issue - State: closed - Opened by Pavelrst almost 4 years ago - 2 comments

#49 - dataloader question when ./train_mixed.sh

Issue - State: closed - Opened by zzl777 almost 4 years ago - 6 comments

#48 - flow_low ?

Issue - State: closed - Opened by shafu0x almost 4 years ago - 1 comment

#47 - chairs checkpoint reproduction

Issue - State: closed - Opened by awaelchli almost 4 years ago - 1 comment

#46 - Not really an issue, per se

Issue - State: closed - Opened by ikanreed almost 4 years ago - 1 comment

#45 - Input Image Pad Mode

Issue - State: closed - Opened by lhao0301 almost 4 years ago - 1 comment

#44 - DAVIS dataset

Issue - State: closed - Opened by robintzeng almost 4 years ago - 2 comments

#43 - Cuda out of memory (when using many frames)

Issue - State: closed - Opened by Wazaki-Ou almost 4 years ago - 2 comments

#42 - A typo in Efficient Correlated layer

Issue - State: closed - Opened by hzhang57 almost 4 years ago - 3 comments

#41 - why concate [x, flow]

Issue - State: closed - Opened by libonwpu almost 4 years ago - 2 comments

#40 - I want to know the flops of model.

Issue - State: closed - Opened by xialulu826 about 4 years ago - 1 comment

#39 - GPU Memory Usage

Issue - State: closed - Opened by lhao0301 about 4 years ago - 4 comments

#38 - What does '2-view' mean in the paper?

Issue - State: closed - Opened by JohnnieXDU about 4 years ago - 1 comment

#37 - is the pretrained models provided same as papers?

Issue - State: closed - Opened by limacv about 4 years ago - 1 comment

#36 - A consultation on the use of RAFT

Issue - State: closed - Opened by Starboy-at-earth about 4 years ago - 1 comment

#35 - Memory issue with one GPU at inference time - process gets killed

Issue - State: closed - Opened by farnaznouraei about 4 years ago - 2 comments

#34 - Nice work, I am curious about model architecture.

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

#33 - How to deal with the DataParallel imbalanced memory usage problem?

Issue - State: closed - Opened by sunnyday2017 about 4 years ago - 4 comments

#32 - failed to load model

Issue - State: closed - Opened by khanhha about 4 years ago - 1 comment

#31 - Why do you use batch normalization in context encoder?

Issue - State: closed - Opened by liuppboy about 4 years ago - 2 comments

#30 - Which of the released model is more suitable for real-world videos

Issue - State: closed - Opened by bo-miao about 4 years ago - 2 comments

#29 - about up mask

Issue - State: closed - Opened by zihuazheng about 4 years ago - 2 comments

#28 - about pretrained models

Issue - State: closed - Opened by JTRNEO about 4 years ago - 1 comment

#27 - Inference Time

Issue - State: closed - Opened by WangDeyao about 4 years ago - 2 comments

#26 - Efficent Implementation

Issue - State: closed - Opened by phongnhhn92 about 4 years ago - 1 comment

#25 - Old flow visualization

Issue - State: closed - Opened by dfrumkin about 4 years ago - 1 comment

#24 - Why scale mask by 0.25 to balance gradients?

Issue - State: closed - Opened by complexfilter about 4 years ago - 1 comment

#23 - Trained Model Load ERROR.

Issue - State: closed - Opened by DwenGu about 4 years ago - 6 comments

#22 - Question about stacking multiple frames

Issue - State: closed - Opened by FedericoVasile1 about 4 years ago - 1 comment

#21 - can not re-implement the results on sintel dataset test split

Issue - State: closed - Opened by Ecalpal about 4 years ago - 8 comments

#20 - Efficient Cost Computation for High Resolution Images

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

#19 - Question about dataset

Issue - State: closed - Opened by llesky about 4 years ago - 2 comments

#18 - small model available?

Issue - State: closed - Opened by complexfilter about 4 years ago - 1 comment

#17 - Data changes on KITTI2015

Issue - State: closed - Opened by Ecalpal about 4 years ago

#16 - Why the edges of the colorized optical flow look jaggy?

Issue - State: closed - Opened by complexfilter about 4 years ago - 7 comments

GitHub / princeton-vl/RAFT issues and pull requests