feat: update ground segmentation evaluator to be compatible with lidarseg dataset #156

nanoshimarobot · 2025-05-28T07:17:14Z

Types of PR

Upgrade of existing features

Description

This PR removes the annotated_rosbag mode from ground_segmentation_evaluator and updates the annotated_pcd mode.

Remove annotated_rosbag
Maintaining this mode requires patching autoware.universe, which has become burdensome and error-prone. So we’re simply discontinuing this mode.
Update annotated_pcd
Until now, segmentation labels were assumed to be written directly into t4_dataset/data/LIDAR_CONCAT/*.pcd.bin.
Since this PR introduced a standard format and location for lidar segmentation labels, annotated_pcd has been re-implemented to follow that spec.
TIER IV internal link

How to review this PR

Test data and scenario

test dataset : TIERIV_INTERNAL_LINK
scenario:

ScenarioFormatVersion: 3.0.0
ScenarioName: ground_segmentation-sample
ScenarioDescription: ground_segmentation-sample
SensorModel: aip_x2_gen2
VehicleModel: j6_gen2
Evaluation:
  UseCaseName: ground_segmentation
  UseCaseFormatVersion: 0.3.0
  Datasets:
    - sample_dataset:
        VehicleId: default
  Conditions:
    Method: annotated_pcd 
    ground_label: 20
    obstacle_label: 7
    accuracy_min: 0.7
    accuracy_max: 1.0
    PassRate: 99.0

Test command

dlr2 simulation run ~/driving_log_replayer_v2 --launch_args rate:=0.1

Test result

{"Condition": {"ground_label": 20, "obstacle_label": 7, "accuracy_min": 0.7, "accuracy_max": 1.0, "PassRate": 99.0}}
{"Result": {"Success": false, "Summary": "NoData"}, "Stamp": {"System": 1748414570.2117581}, "Frame": {}}
{"Result": {"Success": true, "Summary": "Passed: Ground Segmentation (Success): 1 / 1 -> 100.00"}, "Stamp": {"System": 1748414581.7609808, "ROS": 1736225810.2732875}, "Frame": {"GroundSegmentation": {"Result": {"Total": "Success", "Frame": "Success"}, "Info": {"TP": 96739, "FP": 0, "TN": 0, "FN": 6, "Accuracy": 0.9999379812910217, "Precision": 0.999999999999999, "Recall": 0.9999379812910217, "Specificity": 0.0, "F1-score": 0.9999689896339005}}}}
{"Result": {"Success": true, "Summary": "Passed: Ground Segmentation (Success): 2 / 2 -> 100.00"}, "Stamp": {"System": 1748414585.0651531, "ROS": 1736225811.6182482}, "Frame": {"GroundSegmentation": {"Result": {"Total": "Success", "Frame": "Success"}, "Info": {"TP": 96577, "FP": 0, "TN": 0, "FN": 10, "Accuracy": 0.9998964663981684, "Precision": 0.9999999999999989, "Recall": 0.9998964663981684, "Specificity": 0.0, "F1-score": 0.9999482304691433}}}}
{"Result": {"Success": true, "Summary": "Passed: Ground Segmentation (Success): 3 / 3 -> 100.00"}, "Stamp": {"System": 1748414588.832024, "ROS": 1736225815.2883682}, "Frame": {"GroundSegmentation": {"Result": {"Total": "Success", "Frame": "Success"}, "Info": {"TP": 74383, "FP": 0, "TN": 0, "FN": 21799, "Accuracy": 0.7733567611403372, "Precision": 0.9999999999999987, "Recall": 0.7733567611403372, "Specificity": 0.0, "F1-score": 0.8721953506968656}}}}
{"Result": {"Success": true, "Summary": "Passed: Ground Segmentation (Success): 4 / 4 -> 100.00"}, "Stamp": {"System": 1748414591.9241204, "ROS": 1736225818.6884308}, "Frame": {"GroundSegmentation": {"Result": {"Total": "Success", "Frame": "Success"}, "Info": {"TP": 69543, "FP": 0, "TN": 0, "FN": 26267, "Accuracy": 0.7258428139025146, "Precision": 0.9999999999999986, "Recall": 0.7258428139025146, "Specificity": 0.0, "F1-score": 0.8411459120302676}}}}
{"Result": {"Success": true, "Summary": "Passed: Ground Segmentation (Success): 5 / 5 -> 100.00"}, "Stamp": {"System": 1748414594.6325753, "ROS": 1736225822.2882407}, "Frame": {"GroundSegmentation": {"Result": {"Total": "Success", "Frame": "Success"}, "Info": {"TP": 84012, "FP": 0, "TN": 0, "FN": 12888, "Accuracy": 0.8669969040247669, "Precision": 0.9999999999999988, "Recall": 0.8669969040247669, "Specificity": 0.0, "F1-score": 0.928760944497878}}}}
{"Result": {"Success": true, "Summary": "Passed: Ground Segmentation (Success): 6 / 6 -> 100.00"}, "Stamp": {"System": 1748414596.7132423, "ROS": 1736225824.498351}, "Frame": {"GroundSegmentation": {"Result": {"Total": "Success", "Frame": "Success"}, "Info": {"TP": 83723, "FP": 0, "TN": 0, "FN": 12866, "Accuracy": 0.8667964260940678, "Precision": 0.9999999999999988, "Recall": 0.8667964260940678, "Specificity": 0.0, "F1-score": 0.9286459026078692}}}}
{"Result": {"Success": true, "Summary": "Passed: Ground Segmentation (Success): 7 / 7 -> 100.00"}, "Stamp": {"System": 1748414599.0220072, "ROS": 1736225826.5733135}, "Frame": {"GroundSegmentation": {"Result": {"Total": "Success", "Frame": "Success"}, "Info": {"TP": 80775, "FP": 0, "TN": 0, "FN": 14445, "Accuracy": 0.8482986767485813, "Precision": 0.9999999999999988, "Recall": 0.8482986767485813, "Specificity": 0.0, "F1-score": 0.9179238046038771}}}}

Note: If no evaluation results are produced, change the point cloud topic to the following:

dlr2 simulation run ~/driving_log_replayer_v2 --launch_args rate:=0.1 evaluation_target_topic:=<topic_name>

or

ros2 launch driving_log_replayer_v2 driving_log_replayer_v2.launch.py scenario_path:=$HOME/driving_log_replayer_v2/ground_segmentation/scenario.yaml play_rate:=0.1 evaluation_target_topic:=<topic_name>

Others

Copilot

Pull Request Overview

This PR discontinues the annotated_rosbag mode and reimplements annotated_pcd to use the new lidarseg.json spec for segmentation labels.

Drop annotated_rosbag support and related synchronization code
Load separate label files via lidarseg.json in annotated_pcd mode
Update CLOUD_DIM from 6 to 5 for the new point format

Reviewed Changes

Copilot reviewed 2 out of 2 changed files in this pull request and generated no comments.

File	Description
driving_log_replayer_v2/scripts/ground_segmentation_evaluator_node.py	Remove rosbag mode, adjust `annotated_pcd` to parse labels from `lidarseg.json`, and set `CLOUD_DIM=5`.
driving_log_replayer_v2/driving_log_replayer_v2/ground_segmentation.py	Remove `Method` field from `Condition` model to match dropped rosbag mode.

Comments suppressed due to low confidence (3)

driving_log_replayer_v2/scripts/ground_segmentation_evaluator_node.py:22

The import ros2_numpy is no longer used after removing annotated_rosbag_eval_cb; consider removing this unused import.

import ros2_numpy

driving_log_replayer_v2/scripts/ground_segmentation_evaluator_node.py:53

Unconditionally loading annotated PCD mode without validating eval_condition.Method may allow unsupported methods; add a check and raise an error if it's not "annotated_pcd".

sample_data_path = Path(self._t4_dataset_paths[0], "annotation", "sample_data.json")

driving_log_replayer_v2/driving_log_replayer_v2/ground_segmentation.py:27

The Method field was removed from the Condition model but the evaluator still relies on mode configuration; re-add or update the model to match expected scenario parameters.

    Method: Literal["annotated_pcd", "annotated_rosbag"]

badai-nguyen · 2025-05-28T07:30:56Z

@nanoshimarobot
Shall we remove all documantation and others related to annotatated_rosbag such as

driving_log_replayer_v2/sample/ground_segmentation/scenario_annotated_rosbag.yaml

Line 8 in 4b7a336

UseCaseFormatVersion: 0.3.0

driving_log_replayer_v2/docs/use_case/ground_segmentation.en.md

Lines 9 to 15 in 4b7a336

    
           ### annotated_rosbag 
        
           This method involves adding a field to the point cloud data in the bag file to represent semantic labels. 
        
           Synchronize and subscribe to topics before and after ground removal, and evaluate the accuracy by comparing the number of points with ground and non-ground labels. 
        
           In this evaluation framework, the semantic labels are assumed to be recorded in an `INT32` field named `entity_id`.

…cenario

badai-nguyen · 2025-05-29T01:05:51Z

docs/use_case/ground_segmentation.en.md

@@ -6,14 +6,6 @@ Evaluate the performance of the Obstacle Segmentation sub-component in Autoware,

 The Ground Truth data required for evaluation can be provided using the following two methods, and each can be used by changing the `Evaluation.Conditions.Method` of the scenario.


Suggested change

The Ground Truth data required for evaluation can be provided using the following two methods, and each can be used by changing the `Evaluation.Conditions.Method` of the scenario.

The Ground Truth data required for evaluation can be provided using the following method

badai-nguyen · 2025-05-29T01:06:45Z

docs/use_case/ground_segmentation.en.md

-Synchronize and subscribe to topics before and after ground removal, and evaluate the accuracy by comparing the number of points with ground and non-ground labels.
-
-In this evaluation framework, the semantic labels are assumed to be recorded in an `INT32` field named `entity_id`.
-
 ### annotated_pcd

 This method involves adding a field to the point cloud data provided as a dataset (`dataset/data/LIDAR_CONCAT/\*.pcd.bin`) to represent semantic labels.


This's also outdate and need to be updated

badai-nguyen · 2025-05-29T05:13:02Z

@nanoshimarobot could you also put the command you used and the evaluator testing result in description for reference?

nanoshimarobot · 2025-05-29T05:30:18Z

@nanoshimarobot could you also put the command you used and the evaluator testing result in description for reference?

I updated the PR description to include the test command and a sample output.

badai-nguyen · 2025-05-29T08:27:25Z

I also confirmed 'ground_segmentation_evaluator' works with latest lidar semantic segmentation dataset format

{"Condition": {"ground_label": 20, "obstacle_label": 7, "accuracy_min": 0.7, "accuracy_max": 1.0, "PassRate": 99.0}}
{"Result": {"Success": false, "Summary": "NoData"}, "Stamp": {"System": 1748506589.992203}, "Frame": {}}
{"Result": {"Success": true, "Summary": "Passed: Ground Segmentation (Success): 1 / 1 -> 100.00"}, "Stamp": {"System": 1748506605.1093938, "ROS": 1736225810.2802138}, "Frame": {"GroundSegmentation": {"Result": {"Total": "Success", "Frame": "Success"}, "Info": {"TP": 96744, "FP": 0, "TN": 0, "FN": 1, "Accuracy": 0.9999896635485027, "Precision": 0.999999999999999, "Recall": 0.9999896635485027, "Specificity": 0.0, "F1-score": 0.9999948316975401}}}}
{"Result": {"Success": true, "Summary": "Passed: Ground Segmentation (Success): 2 / 2 -> 100.00"}, "Stamp": {"System": 1748506607.7246006, "ROS": 1736225811.70285}, "Frame": {"GroundSegmentation": {"Result": {"Total": "Success", "Frame": "Success"}, "Info": {"TP": 96587, "FP": 0, "TN": 0, "FN": 0, "Accuracy": 0.999999999999999, "Precision": 0.999999999999999, "Recall": 0.999999999999999, "Specificity": 0.0, "F1-score": 0.9999999999499991}}}}
{"Result": {"Success": true, "Summary": "Passed: Ground Segmentation (Success): 3 / 3 -> 100.00"}, "Stamp": {"System": 1748506610.6418622, "ROS": 1736225813.2828286}, "Frame": {"GroundSegmentation": {"Result": {"Total": "Success", "Frame": "Success"}, "Info": {"TP": 68419, "FP": 0, "TN": 0, "FN": 27880, "Accuracy": 0.7104850517658535, "Precision": 0.9999999999999986, "Recall": 0.7104850517658535, "Specificity": 0.0, "F1-score": 0.8307410240046619}}}}
{"Result": {"Success": true, "Summary": "Passed: Ground Segmentation (Success): 4 / 4 -> 100.00"}, "Stamp": {"System": 1748506613.043979, "ROS": 1736225814.480516}, "Frame": {"GroundSegmentation": {"Result": {"Total": "Success", "Frame": "Success"}, "Info": {"TP": 71001, "FP": 0, "TN": 0, "FN": 25074, "Accuracy": 0.7390163934426222, "Precision": 0.9999999999999986, "Recall": 0.7390163934426222, "Specificity": 0.0, "F1-score": 0.8499245851698279}}}}
{"Result": {"Success": true, "Summary": "Passed: Ground Segmentation (Success): 5 / 5 -> 100.00"}, "Stamp": {"System": 1748506615.5768092, "ROS": 1736225815.6578078}, "Frame": {"GroundSegmentation": {"Result": {"Total": "Success", "Frame": "Success"}, "Info": {"TP": 74154, "FP": 0, "TN": 0, "FN": 22028, "Accuracy": 0.7709758582686981, "Precision": 0.9999999999999987, "Recall": 0.7709758582686981, "Specificity": 0.0, "F1-score": 0.8706791282619379}}}}
{"Result": {"Success": true, "Summary": "Passed: Ground Segmentation (Success): 6 / 6 -> 100.00"}, "Stamp": {"System": 1748506618.979621, "ROS": 1736225817.283224}, "Frame": {"GroundSegmentation": {"Result": {"Total": "Success", "Frame": "Success"}, "Info": {"TP": 72536, "FP": 0, "TN": 0, "FN": 23162, "Accuracy": 0.7579677736211825, "Precision": 0.9999999999999986, "Recall": 0.7579677736211825, "Specificity": 0.0, "F1-score": 0.862322717118701}}}}
{"Result": {"Success": true, "Summary": "Passed: Ground Segmentation (Success): 7 / 7 -> 100.00"}, "Stamp": {"System": 1748506621.4693022, "ROS": 1736225818.6404264}, "Frame": {"GroundSegmentation": {"Result": {"Total": "Success", "Frame": "Success"}, "Info": {"TP": 68149, "FP": 0, "TN": 0, "FN": 27661, "Accuracy": 0.711293184427512, "Precision": 0.9999999999999986, "Recall": 0.711293184427512, "Specificity": 0.0, "F1-score": 0.8312931891023684}}}}
{"Result": {"Success": true, "Summary": "Passed: Ground Segmentation (Success): 8 / 8 -> 100.00"}, "Stamp": {"System": 1748506624.3023686, "ROS": 1736225820.2779942}, "Frame": {"GroundSegmentation": {"Result": {"Total": "Success", "Frame": "Success"}, "Info": {"TP": 75367, "FP": 0, "TN": 0, "FN": 20916, "Accuracy": 0.7827653895287842, "Precision": 0.9999999999999987, "Recall": 0.7827653895287842, "Specificity": 0.0, "F1-score": 0.8781473929015133}}}}
{"Result": {"Success": true, "Summary": "Passed: Ground Segmentation (Success): 9 / 9 -> 100.00"}, "Stamp": {"System": 1748506626.4814403, "ROS": 1736225821.3029618}, "Frame": {"GroundSegmentation": {"Result": {"Total": "Success", "Frame": "Success"}, "Info": {"TP": 75938, "FP": 0, "TN": 0, "FN": 20359, "Accuracy": 0.7885811603684426, "Precision": 0.9999999999999987, "Recall": 0.7885811603684426, "Specificity": 0.0, "F1-score": 0.8817952215955432}}}}
{"Result": {"Success": true, "Summary": "Passed: Ground Segmentation (Success): 10 / 10 -> 100.00"}, "Stamp": {"System": 1748506628.569387, "ROS": 1736225822.3955405}, "Frame": {"GroundSegmentation": {"Result": {"Total": "Success", "Frame": "Success"}, "Info": {"TP": 83391, "FP": 0, "TN": 0, "FN": 13509, "Accuracy": 0.8605882352941168, "Precision": 0.9999999999999988, "Recall": 0.8605882352941168, "Specificity": 0.0, "F1-score": 0.9250711349486989}}}}
{"Result": {"Success": true, "Summary": "Passed: Ground Segmentation (Success): 11 / 11 -> 100.00"}, "Stamp": {"System": 1748506630.630671, "ROS": 1736225823.438711}, "Frame": {"GroundSegmentation": {"Result": {"Total": "Success", "Frame": "Success"}, "Info": {"TP": 84998, "FP": 0, "TN": 0, "FN": 11512, "Accuracy": 0.880717024142575, "Precision": 0.9999999999999988, "Recall": 0.880717024142575, "Specificity": 0.0, "F1-score": 0.9365757982621201}}}}
{"Result": {"Success": true, "Summary": "Passed: Ground Segmentation (Success): 12 / 12 -> 100.00"}, "Stamp": {"System": 1748506632.7102108, "ROS": 1736225824.455516}, "Frame": {"GroundSegmentation": {"Result": {"Total": "Success", "Frame": "Success"}, "Info": {"TP": 83125, "FP": 0, "TN": 0, "FN": 13464, "Accuracy": 0.8606052449036631, "Precision": 0.9999999999999988, "Recall": 0.8606052449036631, "Specificity": 0.0, "F1-score": 0.9250809619231922}}}}
{"Result": {"Success": true, "Summary": "Passed: Ground Segmentation (Success): 13 / 13 -> 100.00"}, "Stamp": {"System": 1748506634.834357, "ROS": 1736225825.505582}, "Frame": {"GroundSegmentation": {"Result": {"Total": "Success", "Frame": "Success"}, "Info": {"TP": 83657, "FP": 0, "TN": 0, "FN": 13406, "Accuracy": 0.8618835189516079, "Precision": 0.9999999999999988, "Recall": 0.8618835189516079, "Specificity": 0.0, "F1-score": 0.9258189463867504}}}}
{"Result": {"Success": true, "Summary": "Passed: Ground Segmentation (Success): 14 / 14 -> 100.00"}, "Stamp": {"System": 1748506636.852415, "ROS": 1736225826.56906}, "Frame": {"GroundSegmentation": {"Result": {"Total": "Success", "Frame": "Success"}, "Info": {"TP": 80532, "FP": 0, "TN": 0, "FN": 14688, "Accuracy": 0.8457466918714547, "Precision": 0.9999999999999988, "Recall": 0.8457466918714547, "Specificity": 0.0, "F1-score": 0.9164276935185579}}}}
{"Result": {"Success": true, "Summary": "Passed: Ground Segmentation (Success): 15 / 15 -> 100.00"}, "Stamp": {"System": 1748506638.7968888, "ROS": 1736225827.578199}, "Frame": {"GroundSegmentation": {"Result": {"Total": "Success", "Frame": "Success"}, "Info": {"TP": 76139, "FP": 0, "TN": 0, "FN": 19893, "Accuracy": 0.7928502999000325, "Precision": 0.9999999999999987, "Recall": 0.7928502999000325, "Specificity": 0.0, "F1-score": 0.8844578935564421}}}}
{"Result": {"Success": true, "Summary": "Passed: Ground Segmentation (Success): 16 / 16 -> 100.00"}, "Stamp": {"System": 1748506640.7694693, "ROS": 1736225828.5481339}, "Frame": {"GroundSegmentation": {"Result": {"Total": "Success", "Frame": "Success"}, "Info": {"TP": 73900, "FP": 0, "TN": 0, "FN": 22359, "Accuracy": 0.7677204209476508, "Precision": 0.9999999999999987, "Recall": 0.7677204209476508, "Specificity": 0.0, "F1-score": 0.8685993687764901}}}}

…on launch file

badai-nguyen · 2025-05-29T08:34:23Z

@nanoshimarobot plz fix ci(spell-check) error.

nanoshimarobot · 2025-05-29T08:48:57Z

@nanoshimarobot plz fix ci(spell-check) error.

@badai-nguyen
All the spell-check errors were about lidarseg. Since the dataset directory is named lidarseg, do we really need to change it?

Warning: Unknown word (lidarseg)
docs/use_case/ground_segmentation.en.md:11:165 Unknown word (lidarseg)
docs/use_case/ground_segmentation.ja.md:11:79 Unknown word (lidarseg)
driving_log_replayer_v2/scripts/ground_segmentation_evaluator_node.py:58:78 Unknown word (lidarseg)
sample/ground_segmentation/scenario.ja.yaml:13:113 Unknown word (lidarseg)
sample/ground_segmentation/scenario.yaml:13:1[51](https://github.com/tier4/driving_log_replayer_v2/actions/runs/15319725058/job/43100585476?pr=156#step:3:57) Unknown word (lidarseg)
Warning: Unknown word (lidarseg)
Warning: Unknown word (lidarseg)
Warning: Unknown word (lidarseg)
Warning: Unknown word (lidarseg)
Files checked: 238, Issues found: 5 in 5 files.
Error: 5 spelling issues found in 5 of the 238 files checked.

MasatoSaeki · 2025-06-03T06:30:29Z

I will review it until next week, sorry for delay 🙇

badai-nguyen · 2025-06-17T23:52:04Z

@MasatoSaeki Could you review this PR?

MasatoSaeki · 2025-06-18T04:28:39Z

Sorry, just a minute!

nanoshimarobot added 7 commits March 26, 2025 13:44

wip

d655697

[fix]

789390f

update

e016262

Merge branch 'develop' into feat/eval_ground_segmentation

afe8800

remove annotated_rosbag mode

e103634

Merge branch 'develop' into feat/eval_ground_segmentation

09a88a8

fix

60087aa

Copilot AI review requested due to automatic review settings May 28, 2025 07:17

Copilot AI reviewed May 28, 2025

View reviewed changes

Remove references to annotated_rosbag from documentation and sample s…

369970f

…cenario

badai-nguyen reviewed May 29, 2025

View reviewed changes

fix document

5537345

badai-nguyen requested a review from MasatoSaeki May 29, 2025 08:28

update default value for evaluation_target_topic in ground_segmentati…

4193e82

…on launch file

fix spell

a3a48a2

badai-nguyen changed the title ~~feat: update ground segmentation evaluator~~ feat: update ground segmentation evaluator to be compatible with lidarseg dataset May 30, 2025

nanoshimarobot added 2 commits May 31, 2025 17:48

refactor

fa909cd

fix document

3611990

hayato-m126 mentioned this pull request Jun 3, 2025

feat: delete system defined remap #159

Merged

1 task

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

feat: update ground segmentation evaluator to be compatible with lidarseg dataset #156

feat: update ground segmentation evaluator to be compatible with lidarseg dataset #156

Uh oh!

nanoshimarobot commented May 28, 2025 •

edited

Loading

Uh oh!

Copilot AI left a comment

Uh oh!

badai-nguyen commented May 28, 2025 •

edited

Loading

Uh oh!

badai-nguyen May 29, 2025

Uh oh!

badai-nguyen May 29, 2025

Uh oh!

badai-nguyen commented May 29, 2025 •

edited

Loading

Uh oh!

nanoshimarobot commented May 29, 2025

Uh oh!

badai-nguyen commented May 29, 2025

Uh oh!

badai-nguyen commented May 29, 2025

Uh oh!

nanoshimarobot commented May 29, 2025

Uh oh!

MasatoSaeki commented Jun 3, 2025

Uh oh!

badai-nguyen commented Jun 17, 2025

Uh oh!

MasatoSaeki commented Jun 18, 2025

Uh oh!

Uh oh!

		@@ -6,14 +6,6 @@ Evaluate the performance of the Obstacle Segmentation sub-component in Autoware,

		The Ground Truth data required for evaluation can be provided using the following two methods, and each can be used by changing the `Evaluation.Conditions.Method` of the scenario.

feat: update ground segmentation evaluator to be compatible with lidarseg dataset #156

Are you sure you want to change the base?

feat: update ground segmentation evaluator to be compatible with lidarseg dataset #156

Uh oh!

Conversation

nanoshimarobot commented May 28, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Types of PR

Description

How to review this PR

Test data and scenario

Test command

Test result

Others

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull Request Overview

Reviewed Changes

Uh oh!

badai-nguyen commented May 28, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

badai-nguyen May 29, 2025

Choose a reason for hiding this comment

Uh oh!

badai-nguyen May 29, 2025

Choose a reason for hiding this comment

Uh oh!

badai-nguyen commented May 29, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

nanoshimarobot commented May 29, 2025

Uh oh!

badai-nguyen commented May 29, 2025

Uh oh!

badai-nguyen commented May 29, 2025

Uh oh!

nanoshimarobot commented May 29, 2025

Uh oh!

MasatoSaeki commented Jun 3, 2025

Uh oh!

badai-nguyen commented Jun 17, 2025

Uh oh!

MasatoSaeki commented Jun 18, 2025

Uh oh!

Uh oh!

nanoshimarobot commented May 28, 2025 •

edited

Loading

badai-nguyen commented May 28, 2025 •

edited

Loading

badai-nguyen commented May 29, 2025 •

edited

Loading