docs(README): update poster, video link.

TBD update submit to v2 leaderboard.

Author: Kin-Zhang

.gitignore

@@ -16,4 +16,8 @@ cancel.sh
 
 # slurm log files
 *.err
-*.out
+*.out
+
+# figure
+*.png
+*.pdf

0_preprocess.py

@@ -1,7 +1,7 @@
 """
 # Created: 2023-11-01 17:02
 # Copyright (C) 2023-now, RPL, KTH Royal Institute of Technology
-# Author: Kin ZHANG  (https://kin-zhang.github.io/)
+# Author: Qingwen Zhang  (https://kin-zhang.github.io/)
 #
 # This work is licensed under the terms of the MIT license.
 # For a copy, see <https://opensource.org/licenses/MIT>.

1_train.py

@@ -1,7 +1,7 @@
 """
 # Created: 2023-07-12 19:30
 # Copyright (C) 2023-now, RPL, KTH Royal Institute of Technology
-# Author: Kin ZHANG  (https://kin-zhang.github.io/)
+# Author: Qingwen Zhang  (https://kin-zhang.github.io/)
 #
 # This work is licensed under the terms of the MIT license.
 # For a copy, see <https://opensource.org/licenses/MIT>.
@@ -92,6 +92,7 @@ def main(cfg):
         print("Total Train Dataset Size: ", len(train_dataset))
         print("-"*40+"\n")
 
+    # NOTE(Qingwen): search & check: def training_step(self, batch, batch_idx)
     trainer.fit(model, train_dataloaders = train_loader, val_dataloaders = val_loader, ckpt_path = cfg.checkpoint)
     wandb.finish()
 

2_eval.py

@@ -1,7 +1,7 @@
 """
 # Created: 2023-08-09 10:28
 # Copyright (C) 2023-now, RPL, KTH Royal Institute of Technology
-# Author: Kin ZHANG  (https://kin-zhang.github.io/)
+# Author: Qingwen Zhang  (https://kin-zhang.github.io/)
 #
 # This work is licensed under the terms of the MIT license.
 # For a copy, see <https://opensource.org/licenses/MIT>.

3_vis.py

@@ -1,7 +1,7 @@
 """
 # Created: 2023-12-26 12:41
 # Copyright (C) 2023-now, RPL, KTH Royal Institute of Technology
-# Author: Kin ZHANG  (https://kin-zhang.github.io/)
+# Author: Qingwen Zhang  (https://kin-zhang.github.io/)
 #
 # This work is licensed under the terms of the MIT license.
 # For a copy, see <https://opensource.org/licenses/MIT>.
@@ -35,8 +35,8 @@ def main(cfg):
     mymodel = ModelWrapper.load_from_checkpoint(cfg.checkpoint, cfg=cfg, eval=True)
 
     wandb_logger = WandbLogger(save_dir=output_dir,
-                               entity="hdmaptest",
-                               project=f"sceneflow-eval", 
+                               entity="kth-rpl",
+                               project=f"deflow-eval", 
                                name=f"{cfg.output}",
                                offline=(cfg.wandb_mode == "offline"))
 

README.md

@@ -1,14 +1,15 @@
-DeFlow 
+DeFlow: Decoder of Scene Flow Network in Autonomous Driving
 ---
 
-[![arXiv](https://img.shields.io/badge/arXiv-2401.16122-b31b1b.svg)](https://arxiv.org/abs/2401.16122) [![PWC](https://img.shields.io/endpoint.svg?url=https://paperswithcode.com/badge/deflow-decoder-of-scene-flow-network-in/scene-flow-estimation-on-argoverse-2)](https://paperswithcode.com/sota/scene-flow-estimation-on-argoverse-2?p=deflow-decoder-of-scene-flow-network-in) [poster coming soon] [video coming soon]
+[![arXiv](https://img.shields.io/badge/arXiv-2401.16122-b31b1b?logo=arxiv&logoColor=white)](https://arxiv.org/abs/2401.16122) 
+[![PWC](https://img.shields.io/endpoint.svg?url=https://paperswithcode.com/badge/deflow-decoder-of-scene-flow-network-in/scene-flow-estimation-on-argoverse-2)](https://paperswithcode.com/sota/scene-flow-estimation-on-argoverse-2?p=deflow-decoder-of-scene-flow-network-in) 
+[![poster](https://img.shields.io/badge/Poster-6495ed?style=flat&logo=Shotcut&logoColor=wihte)](https://hkustconnect-my.sharepoint.com/:b:/g/personal/qzhangcb_connect_ust_hk/EXP_uXYmm_tItTWc8MafXHoB-1dVrMnvF1-lCzU1PXAvqQ?e=2FPfBS) 
+[![video](https://img.shields.io/badge/video-YouTube-FF0000?logo=youtube&logoColor=white)](https://youtu.be/bZ4uUv0nDa0)
 
 Will present in ICRA'24.
 
 Task: Scene Flow Estimation in Autonomous Driving. Pre-trained weights for models are available in [Onedrive link](https://hkustconnect-my.sharepoint.com/:f:/g/personal/qzhangcb_connect_ust_hk/Et85xv7IGMRKgqrVeJEVkMoB_vxlcXk6OZUyiPjd4AArIg?e=lqRGhx). Check usage in [2. Evaluation](#2-evaluation) or [3. Visualization](#3-visualization). 
 
-https://github.com/KTH-RPL/DeFlow/assets/35365764/15581af1-3066-4865-bf72-1242a478b938
-
 **Scripts** quick view in our scripts:
 
 - `0_preprocess.py` : pre-process data before training to speed up the whole training time.
@@ -111,7 +112,9 @@ python tests/scene_flow.py --flow_mode='flow_est' --data_dir=/home/kin/data/av2/
 
 Note: ego_motion already compensated, so the visualization is more clear.
 
-![](assets/docs/vis_res.png)
+<!-- ![](assets/docs/vis_res.png) -->
+
+https://github.com/KTH-RPL/DeFlow/assets/35365764/15581af1-3066-4865-bf72-1242a478b938
 
 ## Cite & Acknowledgements
 
@@ -124,6 +127,8 @@ Note: ego_motion already compensated, so the visualization is more clear.
 }
 ```
 
-This implementation is based on codes from several repositories. Thanks to these authors who kindly open-sourcing their work to the community. Please see our paper reference part to get more information. Thanks to Kyle Vedder (ZeroFlow) who kindly discussed their results with us and HKUST Ramlab's member: Jin Wu who gave constructive comments on this work. The computations were enabled by the supercomputing resource Berzelius provided by National Supercomputer Centre at Linköping University and the Knut and Alice Wallenberg Foundation, Sweden.
+This implementation is based on codes from several repositories. Thanks to these authors who kindly open-sourcing their work to the community. Please see our paper reference part to get more information. 
+Thanks to [Kyle Vedder (ZeroFlow)](https://github.com/kylevedder) who kindly discussed their results with us and HKUST Ramlab's member: Jin Wu who gave constructive comments on this work. 
+The computations were enabled by the supercomputing resource Berzelius provided by National Supercomputer Centre at Linköping University and the Knut and Alice Wallenberg Foundation, Sweden.
 
 ❤️: [ZeroFlow](https://github.com/kylevedder/zeroflow), [NSFP](https://github.com/Lilac-Lee/Neural_Scene_Flow_Prior), [FastNSF](https://github.com/Lilac-Lee/FastNSF). Others good code style and tools: [forecast-mae](https://github.com/jchengai/forecast-mae), [kiss-icp](https://github.com/PRBonn/kiss-icp)

scripts/network/dataloader.py

@@ -1,7 +1,7 @@
 """
 # Created: 2023-11-04 15:52
 # Copyright (C) 2023-now, RPL, KTH Royal Institute of Technology
-# Author: Kin ZHANG  (https://kin-zhang.github.io/)
+# Author: Qingwen Zhang  (https://kin-zhang.github.io/)
 #
 # This work is licensed under the terms of the MIT license.
 # For a copy, see <https://opensource.org/licenses/MIT>.

scripts/network/loss_func.py

@@ -1,7 +1,7 @@
 """
 # Created: 2023-07-17 00:00
 # Copyright (C) 2023-now, RPL, KTH Royal Institute of Technology
-# Author: Kin ZHANG  (https://kin-zhang.github.io/)
+# Author: Qingwen Zhang  (https://kin-zhang.github.io/)
 #
 # This work is licensed under the terms of the MIT license.
 # For a copy, see <https://opensource.org/licenses/MIT>.

scripts/network/models/deflow.py

@@ -2,7 +2,7 @@
 """
 # Created: 2023-07-18 15:08
 # Copyright (C) 2023-now, RPL, KTH Royal Institute of Technology
-# Author: Kin ZHANG  (https://kin-zhang.github.io/)
+# Author: Qingwen Zhang  (https://kin-zhang.github.io/)
 #
 # This work is licensed under the terms of the MIT license.
 # For a copy, see <https://opensource.org/licenses/MIT>.

scripts/pl_model.py

@@ -2,7 +2,7 @@
 
 # Created: 2023-11-05 10:00
 # Copyright (C) 2023-now, RPL, KTH Royal Institute of Technology
-# Author: Kin ZHANG  (https://kin-zhang.github.io/)
+# Author: Qingwen Zhang  (https://kin-zhang.github.io/)
 #
 # This work is licensed under the terms of the MIT license.
 # For a copy, see <https://opensource.org/licenses/MIT>.