Feature/visualization　（visualize_bbox functionに修正が必要なので） #32

AtomScott · 2022-09-18T06:10:14Z

No description provided.

soccertrack/utils/DataFrame/bboxdataframe.py

AtomScott · 2022-09-18T06:32:33Z

soccertrack/utils/DataFrame/bboxdataframe.py

+        BB_WIDTH_INDEX = 2
+        BB_HEIGHT_INDEX = 3
+
+        unique_team_ids = self.columns.get_level_values("TeamID").unique()


ここのfor loopは

frame_df = self.loc[self.index==frame_idx] bboxes = frame_df.to_list()

にしてから

for bbox in bboxes: x, y, x2, y2 = bbox[:4] frame = add ...

的な感じで出力するのがシンプルだと思いますが、修正可能でしょうか？

AtomScott · 2022-09-18T06:44:19Z

soccertrack/utils/DataFrame/bboxdataframe.py

        """
+


なんとなく、この関数はほぼbbdf.stack(('TeamID', 'PlayerID')).reset_index().valuesで簡略化できると思います。
pandasのstack/unstackあたりのドキュメンテーションが参考になります。
https://pandas.pydata.org/docs/user_guide/reshaping.html#reshaping-by-stacking-and-unstacking

（ほぼなので列の入れ替えとか、obj_idの作成が必要）

AtomScott · 2022-09-18T06:46:12Z

soccertrack/utils/DataFrame/bboxdataframe.py

+                bb_right(float) : Bounding box right coordinate
+                bb_bottom(float) : Bounding box bottom coordinate
+                conf(float) : Confidence score
+                class_id(int) : Class ID. In the normal case, follow the value of the coco image dataset.


cocoのidであることはどこで使ってる？

これ評価指標を算出するときに使ってます。（可視化のときには使ってない）

AtomScott · 2022-09-18T06:47:42Z

soccertrack/utils/DataFrame/bboxdataframe.py

+                bb_bottom(float) : Bounding box bottom coordinate
+                conf(float) : Confidence score
+                class_id(int) : Class ID. In the normal case, follow the value of the coco image dataset.
+                obj_id(int) : Object ID. This id is a unique integer value for each track


obj_idがplayer_idとteam_idから作られるなら明記しといたほうがいいような気します。

AtomScott · 2022-09-18T06:48:48Z

soccertrack/utils/DataFrame/bboxdataframe.py

        """Convert a dataframe column to a 2-dim list for evaluation of object detection.

        Args:
-            df (pd.DataFrame): Dataframe
+            df (BBoxDataFrame): BBoxDataFrame
            xywh (bool): If True, convert to x1y1x2y2 format. Defaults to True.


これあってる？

AtomScott · 2022-09-18T06:49:38Z

soccertrack/utils/DataFrame/bboxdataframe.py

+            yield img_
+
+
+def add(


add_bbox_to_frameの方が適切？

変更しました

AtomScott · 2022-09-18T06:52:40Z

soccertrack/utils/DataFrame/bboxdataframe.py

+    colors = [_rgb_to_bgr(item) for item in _COLOR_NAME_TO_RGB[color]]
+    color_value, _ = colors
+
+    image = cv2.rectangle(image, (left, top), (right, bottom), color, 2)


cv２.rectangleにのcolorってstrを受け付けないようながするのでなおしてください（これで動作する？）

一応strは受け付けるみたいです。
http://opencv.jp/opencv-2svn/cpp/drawing_functions.html#:~:text=*it%3B-,cv%3A%3Arectangle,-%C2%B6

AtomScott · 2022-09-18T06:56:25Z

soccertrack/utils/DataFrame/bboxdataframe.py

+BB_WIDTH_INDEX = 3
+BB_HEIGHT_INDEX = 4
+
+color_list = []


ランダムに色を生成する関数を作ってもいいかもね
参考になると思うのでhttps://stackoverflow.com/questions/28999287/generate-random-colors-rgb

AtomScott · 2022-09-18T07:02:48Z

コード見ました。コメントした部分が直されていてよくなったと思います。

pandasを活かしてよりシンプルな方法でdataframeを扱えるところがいくつかありそうだったのでpandasのuser guideなどを参考にしてみてください。

https://pandas.pydata.org/docs/user_guide/reshaping.html#reshaping-by-stacking-and-unstacking
https://pandas.pydata.org/docs/user_guide/advanced.html
https://pandas.pydata.org/docs/user_guide/groupby.html

あと、動作確認をするのに作ったコードをテストとして書き直して、SoccerTrack/tests/test_dataframe.pyに入れてくれると助かります。

…rack into feature/visualization

review-notebook-app · 2022-09-25T18:48:02Z

Check out this pull request on

See visual diffs & provide feedback on Jupyter Notebooks.

Powered by ReviewNB

IkumaUchida · 2022-09-25T18:51:08Z

soccertrack/dataframe/bboxdataframe.py

 from ..utils import MovieIterator
 from .base import SoccerTrackMixin
-from ..logging import logger
+


to_listメソッドでPerformanceWarningが発生して、出力に影響が出ているわけではないので警告を消すようなメソッcontainerを追加しました。

pandasの開発者も上記メソッドで避けても良い（？）と言っているみたいなので参考にしています。
twopirllc/pandas-ta#340

どんなwarning だった？

PerformanceWarning: DataFrame is highly fragmented. This is usually the result of calling `frame.insert` many times, which has poor performance. Consider using pd.concat instead. To get a de-fragmented frame, use `newframe = frame.copy()`

と一緒？

コードを治すためのヒントが含まれている場合もあります。

IkumaUchida · 2022-09-25T18:53:30Z

soccertrack/dataframe/bboxdataframe.py

+BB_BOTTOM_INDEX = 4
+BB_WIDTH_INDEX = 3
+BB_HEIGHT_INDEX = 4
+LABEL_INDEX = 7

 _COLOR_NAMES = list(_COLOR_NAME_TO_RGB)




色をランダムに生成する部分は未実装です。すみません。

このままでOKです。

IkumaUchida · 2022-09-25T19:01:42Z

file.pyにて、load_motとload_dfの出力が揃っていなかったので修正しました。具体的には、BBoxDataFrameが定めるAttributeを生成できるように修正しています。

bb_left
bb_top
bb_width
bb_height
conf

AtomScott

全体的にコードの質とドキュメンテーションの充実度がよくなってきていると思います。

コメントしたところの確認お願いします。

AtomScott · 2022-09-26T00:05:04Z

soccertrack/dataframe/bboxdataframe.py

 from ..utils import MovieIterator
 from .base import SoccerTrackMixin
-from ..logging import logger
+


どんなwarning だった？

PerformanceWarning: DataFrame is highly fragmented. This is usually the result of calling `frame.insert` many times, which has poor performance. Consider using pd.concat instead. To get a de-fragmented frame, use `newframe = frame.copy()`

と一緒？

コードを治すためのヒントが含まれている場合もあります。

AtomScott · 2022-09-26T00:09:00Z

soccertrack/dataframe/bboxdataframe.py

@@ -78,74 +111,105 @@ def _constructor(self: pd.DataFrame) -> Type[BBoxDataFrame]:

        # df: pd.DataFrame

-    def to_list(self: BBoxDataFrame, xywh: bool = True) -> list[list[float]]:
+    def get_id(self: BBoxDataFrame) -> list[tuple[Any, Any]]:


get_idが何を指しているかが不明なので、get_player_ids, get_teams_idsに分けた方がいい気がします。もし選手のattributesに対してループをかけたいのであれば、soccertrack mixinで実装したiter_playersを使ってください。

AtomScott · 2022-09-26T00:10:30Z

soccertrack/dataframe/bboxdataframe.py

+        Returns:
+            id_list(list[tuple[Any, Any]]): List of team id and player id.
+        """
+        id_list = []


この関数、
self.columns.levels[0]
と
self.columns.levels[1]
で代用できませんか？

AtomScott · 2022-09-26T00:21:02Z

soccertrack/dataframe/bboxdataframe.py

+        # append_object_id
+        obj_id = 0
+
+        for id in sorted(set(id_list)):


SoccerTrackMixinのto_long_df()を使ってからclass_id, team_player_idなどをforループ使わずに計算できそうな気もします。

bbdf.iter_players()

df = df.to_long_df() df['class_id'] = f(team_id, player_id)から計算される df['obj_id'] = {team_id}_{player_id}?

AtomScott · 2022-09-26T00:21:36Z

soccertrack/dataframe/bboxdataframe.py

        for bbox in bbox_2dim_list:
+            bbox[FRAME_INDEX] = bbox[FRAME_INDEX] - frame_idx_start  # start from 0
            if xywh:


xywhってどんな状況で使うんだっけ？

to_listに渡されるbboxの列がxywhの場合は、xyx2y2に変更するような形を意図したはず。
けどsoccertrackmixinを使うならあまり必要がないかも知れない

AtomScott · 2022-09-26T00:24:57Z

soccertrack/dataframe/bboxdataframe.py

+            )
+            frame = add(
+                frame, int(x), int(y), int(x2), int(y2), label=label, color=color
+            )
        return frame

    def visualize_bbox(self, video_path: str) -> Iterator[np.ndarray]:


分かりにくいのでvisualize_framesとかに変更しちゃいましょう。

visualize_framesに変更しました。

AtomScott · 2022-09-26T00:26:41Z

soccertrack/dataframe/bboxdataframe.py

@@ -202,10 +251,18 @@ def visualize_bbox(self, video_path: str) -> Iterator[np.ndarray]:
        Returns:
            None
        """
+
        movie_iterator = MovieIterator(video_path)


この関数、変更前の状態に戻されてそうな予感（コメント・不要なfor-loop等）

pullが間違ったのかも知れません。
major refactorのコードに書き換えます。

AtomScott · 2022-09-26T00:28:48Z

soccertrack/io/file.py

@@ -357,18 +357,18 @@ def load_mot(filename: _pathlike) -> GPSDataFrame:

    df_list = []
    for playerid, group in groups:
-        group['conf'] = 1.0
-        group['class_id'] = int(0) # TODO: classid of person
+        group["conf"] = 1.0


conf=1にするとロードしたデータも上書きしちゃいそうです。

このままだと確かにそうですね、
confの列を引き続きBBoxdataframeに入れるのならば、
推論時の出力と、正解データの出力でconfに渡す値を切り替えられるようにしたいかもです
(load_dfの引数にgt=True/Falseを追加するイメージ)

AtomScott · 2022-09-26T00:30:59Z

soccertrack/io/file.py

+        player_id = column[1]
+        id_list.append((team_id, player_id))
+
+    for id in sorted(set(id_list)):


conf=1にするとロードしたデータも上書きしちゃいそうです。

AtomScott · 2022-09-26T00:33:11Z

soccertrack/utils/utils.py

@@ -7,11 +7,13 @@



このファイルはコミットミス?

ごめんなさいなにも触れずにコミットしてました。
230行目のloopですが、tqdmメソッドを入れることによってエラーが発生しました。その為tqdmを一時的に外しています。

ただそのエラー内容を保存していなかったので、後でエラーを再現して再度コメントします。

utilsのmake_videoメソッドの231行目で以下のエラーが出ます
該当コード

for frame in tqdm(frames, desc=f"Writing video", level="INFO"):

エラー内容

/home/guest/.local/lib/python3.9/site-packages/rich/live.py:229: UserWarning: install "ipywidgets" for Jupyter support warnings.warn('install "ipywidgets" for Jupyter support') None Exception in thread Thread-6: Traceback (most recent call last): File "/opt/anaconda3/envs/soccertrack/lib/python3.9/threading.py", line 973, in _bootstrap_inner self.run() File "/home/guest/.local/lib/python3.9/site-packages/rich/live.py", line 32, in run self.live.refresh() File "/home/guest/.local/lib/python3.9/site-packages/rich/live.py", line 221, in refresh self._live_render.set_renderable(self.renderable) File "/home/guest/.local/lib/python3.9/site-packages/rich/live.py", line 203, in renderable renderable = self.get_renderable() File "/home/guest/.local/lib/python3.9/site-packages/rich/live.py", line 98, in get_renderable self._get_renderable() File "/home/guest/.local/lib/python3.9/site-packages/rich/progress.py", line 1539, in get_renderable /home/guest/dev_repo/SoccerTrack/soccertrack/logging.py:117: TqdmExperimentalWarning: rich is experimental/alpha return tqdm(*args, **kwargs) renderable = Group(*self.get_renderables()) File "/home/guest/.local/lib/python3.9/site-packages/rich/progress.py", line 1544, in get_renderables table = self.make_tasks_table(self.tasks) File "/home/guest/.local/lib/python3.9/site-packages/rich/progress.py", line 1568, in make_tasks_table table.add_row( File "/home/guest/.local/lib/python3.9/site-packages/rich/progress.py", line 1573, in <genexpr> else column(task) File "/home/guest/.local/lib/python3.9/site-packages/rich/progress.py", line 524, in __call__ renderable = self.render(task) File "/opt/anaconda3/envs/soccertrack/lib/python3.9/site-packages/tqdm/rich.py", line 34, in render total = int(task.total) TypeError: int() argument must be a string, a bytes-like object or a number, not 'NoneType' Output exceeds the [size limit](command:workbench.action.openSettings?[). Open the full output data [in a text editor](command:workbench.action.openLargeOutput?733c860c-793e-437c-bf34-d2a280d19f0c) --------------------------------------------------------------------------- TypeError Traceback (most recent call last) Cell In [12], line 7 5 save_dir = './' 6 save_path = os.path.join(save_dir, 'bbox_FISH.mp4') ----> 7 make_video(img_, save_path) File ~/dev_repo/SoccerTrack/soccertrack/utils/utils.py:231, in make_video(frames, outpath, vcodec, preset, crf, ss, t, c, height, width, input_framerate, logging) 226 writer = WriteGear( 227 output_filename=outpath, compression_mode=True, logging=logging, **output_params 228 ) 230 # loop over --> 231 for frame in tqdm(frames, desc=f"Writing video", level="INFO"): 232 233 # simulating RGB frame for example 234 frame_rgb = frame[:, :, ::-1] 236 # writing RGB frame to writer File ~/dev_repo/SoccerTrack/soccertrack/logging.py:117, in tqdm(level, *args, **kwargs) 115 enable = logger.level(LOG_LEVEL).no <= logger.level(level.upper()).no 116 kwargs.update({"disable": not enable}) --> 117 return tqdm(*args, **kwargs) File /opt/anaconda3/envs/soccertrack/lib/python3.9/site-packages/tqdm/rich.py:117, in tqdm_rich.__init__(self, *args, **kwargs) 115 self._prog = Progress(*progress, **options) ... 38 ["", "K", "M", "G", "T", "P", "E", "Z", "Y"], 39 self.unit_divisor, 40 ) TypeError: int() argument must be a string, a bytes-like object or a number, not 'NoneType'

intにNoneTypeが渡されると出るけど、どの部分に該当しているのかがわからず、、

またlogging.pyのtqdmメソッドで使われるtqdm.richに原因があるのかと思って、ドキュメンテーションからソースコードを見ようとしたけど、リンクが切れていたので確認出来ていないです。

また、 from tqdm.rich import tqdmを from tqdm import tqdmに変更すると、動作はするみたいで、以下のような可視化はされます。

いずれにせよ、今回PRのスコープ外な気がしますが、動作する形で修正箇所をコミットします。（utils.pyとlogging.pyをコミット予定）

これ1~2行目の

/home/guest/.local/lib/python3.9/site-packages/rich/live.py:229: UserWarning: install "ipywidgets" for Jupyter support warnings.warn('install "ipywidgets" for Jupyter support') None

が問題だったりする？ pip install ipywidgets で解決するかも

AtomScott · 2022-09-26T07:16:55Z

soccertrack/dataframe/bboxdataframe.py

-
+        self.index: pd.MultiIndex
+
+        if self.index[0] != 0:


元のファイルにどうにかしたい
229~230消去でOK

次のコミットでは消します。

AtomScott · 2022-09-26T07:19:32Z

soccertrack/dataframe/bboxdataframe.py

+            logger.debug(
+                f"x:{x}, y:{y}, x2:{x2}, y2:{y2}, label:{label}, color:{color}"
+            )
+            frame = add(


add変更が反映されていない模様(?)

add_bbox_to_frameという名前に変更しました。

AtomScott

色分けは急ぎじゃないので、PRをマージするところを今日中に行う
・utils.pyをPRから外す
・File.pyはそのまま
・bboxdataframeはよう変更

AtomScott · 2022-09-26T07:25:18Z

soccertrack/dataframe/bboxdataframe.py

+BB_BOTTOM_INDEX = 4
+BB_WIDTH_INDEX = 3
+BB_HEIGHT_INDEX = 4
+LABEL_INDEX = 7

 _COLOR_NAMES = list(_COLOR_NAME_TO_RGB)




このままでOKです。

AtomScott · 2022-09-26T07:29:05Z

soccertrack/dataframe/bboxdataframe.py

+        # append_object_id
+        obj_id = 0
+
+        for id in sorted(set(id_list)):


bbdf.iter_players()

AtomScott · 2022-09-26T07:40:17Z

soccertrack/dataframe/bboxdataframe.py

+        # append_object_id
+        obj_id = 0
+
+        for id in sorted(set(id_list)):


df = df.to_long_df() df['class_id'] = f(team_id, player_id)から計算される df['obj_id'] = {team_id}_{player_id}?

AtomScott · 2022-09-26T07:43:21Z

soccertrack/dataframe/bboxdataframe.py

+        if self.index[0] != 0:
+            self.index = self.index - self.index[0]
+        frame_df = self.loc[self.index == frame_idx].copy()
+        bboxes = frame_df.to_list()


for index, group in frame_df.iter_players(): x, y, x2, y2 = group['bbox', ... label = index[0] + '_' + index[1]

AtomScott · 2022-09-26T07:46:05Z

soccertrack/dataframe/bboxdataframe.py

+            id_list.append((team_id, player_id))
+        return id_list
+
+    def to_list(self: BBoxDataFrame, xywh: bool = True) -> Any:


to_listは外して, SoccerTrackMixinの関数を使う

to_listのうちbboxdfに列(obj_idなど)を追加する箇所が別のメソッド()として持っておいてもよさそうで、それ以外はSoccerTrackMixinの関数を使う

列を追加する機能は便利そうなので、SoccerTrackMixinにadd_attribute的な関数を加えてもいいと思いました

AtomScott · 2022-09-27T06:32:09Z

OK!

* add poetry conf * fix sphinx config path * rename sphinx config * update docs * update project requirements * add poetry reqs command * add logo * hide sidebar name * Update README.md * add citation * change furo colours * add temp * Update README.md * Update video paths * squash later * Update pyproject.toml * docs: 📝 bulk update documentations * build: ⬆️ update dependencies * add iou.ipynb(not completed) (#24) * add iou.ipynb(not completed) * add iou function * add ap&map * fix calc_ap and calc_map * fix evaluation function * add metrics.py * add scripts into metrics.py * move metrics.py to sandbox * fix metrics.py * add mots eval code(no fix from Trackeval) * refactor: ♻️ re-organize code structure * test: 🧪 add test structure * add iou_score tests * fix: 🐛 fix iou_score related methods * Add Google-style docstring to object detection metrics * add prospector automation * Update prospector.yml Co-authored-by: Atom Scott <[email protected]> * Delete sandbox directory * Delete papers directory * Add prospector workflow (#26) * Update and rename prospector.yml to lint.yml * Update lint.yml * update requirements * update requirements * try new config * add paper preview image * add paper preview image * docs: 📝 Update dataset install instructions * docs: 📝 add Kaggle and PWC badges * Fix table * Add video links for github * refactor: ♻️ pass prospector * add updates bbox dataframe handling * update document sections (sidebar) * Function/mot eval (#27) * add iou.ipynb(not completed) * add iou function * add ap&map * fix calc_ap and calc_map * fix evaluation function * add metrics.py * add scripts into metrics.py * move metrics.py to sandbox * fix metrics.py * add mots eval code(no fix from Trackeval) * add hota metrics(Codes not yet organized) * add hota metrics(Codes not yet organized) * add mot_evaluation_metrics * delete sandbox * fix mot metrics * delete sandbox (re commit) * delete run_eval.py and metrics.py * Apply the evaluation index to bbox_df * Fix the path of the library to import * remove unnecessary comments Co-authored-by: Atom Scott <[email protected]> * add chapter 1 on intro to soccertrack * add chapter 2 & 3 on intro to soccertrack * feat: ✨ better logging tools * style: 🎨 * fix: 🐛 fix inspect bug * feat: ✨ add automatic dataset downloader * add small sample csv file * docs: 📝 Add `Dataset Preparation` tutorial * docs: 📝 make dataframes viewable * add sample tracking data * fix: 🐛 add custom prompt & confirm functions * build: ➕ add kaggle dependency * chore: 🧑‍💻 add make command for lint * feat: ✨ add styling function to make dataframes visible in docs * add kaggle import * docs: 📝 update docs * add acknowledgements * (bugfix) copy notebooks to docs on rtd build (#31) * update docs * copy notebooks to docs on rtd build * feature/visualization (#29) * Added the function to visualize the bbox in one frame. * Added video-enabled visualization function * Add Visualization Functions * Apply prospector to bboxdataframe * update notebooks * remove old notebook * Organize soccertrack content * update index * Add star history * feat: ✨ add useful iter accessors * Rename GPSDataFrame to CoordinatesDataFrame * Feature/visualization　（visualize_bbox functionに修正が必要なので） (#32) * Added the function to visualize the bbox in one frame. * Added video-enabled visualization function * Add Visualization Functions * Apply prospector to bboxdataframe * fix visualize_bbox function * fix visualization * fix visualization functions * del index sort of visualize_frame in bboxdataframe * remove utils.py from PR * fix utils and logging * fix utils * fix utils * fix utils * fix utils * fix logging * fix logging * fix logging * fix logging * add comments to code * delete comments * fix visualization functions Co-authored-by: IkumaUchida <[email protected]> * update dataframe * typo * feat: ✨ Easier way to access downloaded dataset (#36) * small fixes * Feature/visualization (#37) * Added the function to visualize the bbox in one frame. * Added video-enabled visualization function * Add Visualization Functions * Apply prospector to bboxdataframe * fix visualize_bbox function * fix visualization * fix visualization functions * del index sort of visualize_frame in bboxdataframe * remove utils.py from PR * fix utils and logging * fix utils * fix utils * fix utils * fix utils * fix logging * fix logging * fix logging * fix logging * add comments to code * delete comments * fix visualization functions * add get_frame Co-authored-by: IkumaUchida <[email protected]> * add pix_fmt option to make_video * update bboxdataframe.py * fix fps rate bug * Update README.md * Update README.md * ignore external directory * add to_yolov5_format function * fix mapping when arg is None * don't ignore .gitkeep! * add tutorials to train yolov5 model * add script to build dataset in YOLOv5 format * bug fix for logger * fix default mapping for yolov5 * read dataframe from numpy (statsbomb) * add tracking notebook for deep sort * add issue template * Docs/get started (#39) * update docs * copy notebooks to docs on rtd build * Add .deepsource.toml * add bbdf2ndjson and upload function🔼 (#38) * add bbdf2ndjson and upload function🔼 * fix uploader * fix file.py * Update .deepsource.toml * Fix dangerous default argument (#40) Co-authored-by: deepsource-autofix[bot] <62050782+deepsource-autofix[bot]@users.noreply.github.com> * update logger tests * update tests for ap_score * clean up commented code * add tests and update map_score * bug fix * add get_frame function * fix soccertrack.logging to soccertrack.logger * adding notebook for yolov5 * adding notebook for appearance feature extraction * Rename appearance_feature_extraction_with_arcface.ipynb to appearance_feature_extraction_with_resnet18.ipynb * Function/detection eval (#41) * fix so that the test passes. * delete comment out * fix object_detection_metrics * Merge PR request * Fix iou_scores for when bbox is out of range * Add first draft of training with a basic motion model * Add iter_frames() * Add to_list_of_tuples_format * add tests for get_git_root() * fix imports from sandbox * Pull Request: Fix of Mot evaluation (#43) * Add preprocessing functions for mot metrics * fix mot eval script * add mot eval test(only case 1) * fix tracking evaluation notebook(add description) * add details of tracking metrics * reflect docs change and test acceleration * Change index for bbox info to a global variable * Fix metrics so that tests pass * replace np.float with generic float * replace np.int with generic int * add idf1, DetA, AssA examples * Fix evaluation script to handle missing track data * add drop option to iteration methods * use iter_players instead of groupby Co-authored-by: Atom Scott <[email protected]> * black * update poetry * black * black * trying to merge a stash pop after too long... * merge new fixes * merge * add detection model * add image feature extractor model * add tracking model * cleanup notebooks * update poetry * Update poetry from 1.2.0b1 to 1.3.1 * add draw function * Create LICENSE * fix: incorrect type casting * improve performance speed of `preprocess_for_mot_eval` * Remove Codacy badge * Update README.md * Update .deepsource.toml * Format code with black and isort (#45) This commit fixes the style issues introduced in 34fe885 according to the output from black and isort. Details: https://deepsource.io/gh/AtomScott/SoccerTrack/transform/22f41986-1297-4a37-8e40-e2e595cc9c46/ Co-authored-by: deepsource-autofix[bot] <62050782+deepsource-autofix[bot]@users.noreply.github.com> * Format code with black and isort (#44) This commit fixes the style issues introduced in 34fe885 according to the output from black and isort. Details: https://deepsource.io/gh/AtomScott/SoccerTrack/transform/ba1386e0-f471-4159-8620-a000d16e741d/ Co-authored-by: deepsource-autofix[bot] <62050782+deepsource-autofix[bot]@users.noreply.github.com> * Update the docs! (#46) * add Documentation section * bulk update of the docs * ignore autoapi * add sphinx-autoapi * add sphinx-autoapi * wasn't necessary * add checks for detections and trackers * add single object tracker * add matching functions * add CostMatrixMetrics * fix circular import * refactor multi_object_tracker * refactor imports * add notebook tutorials * Format code with black and isort (#47) This commit fixes the style issues introduced in fe62d8b according to the output from black and isort. Details: https://deepsource.io/gh/AtomScott/SoccerTrack/transform/55d99477-e19f-4c8c-9022-a5938920150b/ Co-authored-by: deepsource-autofix[bot] <62050782+deepsource-autofix[bot]@users.noreply.github.com> * Format code with black and isort (#48) This commit fixes the style issues introduced in fe62d8b according to the output from black and isort. Details: https://deepsource.io/gh/AtomScott/SoccerTrack/transform/526cd8e9-59c0-4fd5-808c-9da21d89d84e/ Co-authored-by: deepsource-autofix[bot] <62050782+deepsource-autofix[bot]@users.noreply.github.com> * add a working muli object tracker * add EuclideanCMM * add EuclideanCMM * hotfix: skip sim_score compute when length is zero * hotfix for linear_sum_assignment_with_inf * add KalmanTracker prototype * Format code with black and isort (#49) This commit fixes the style issues introduced in 89b26b0 according to the output from black and isort. Details: https://deepsource.io/gh/AtomScott/SoccerTrack/transform/503ec624-ab31-47fb-8df2-dca9a2c533c7/ Co-authored-by: deepsource-autofix[bot] <62050782+deepsource-autofix[bot]@users.noreply.github.com> * Format code with black and isort (#50) This commit fixes the style issues introduced in 89b26b0 according to the output from black and isort. Details: https://deepsource.io/gh/AtomScott/SoccerTrack/transform/5ac4c7da-4986-4b96-bba8-3d1305e2e2ea/ Co-authored-by: deepsource-autofix[bot] <62050782+deepsource-autofix[bot]@users.noreply.github.com> * ignore tests/outputs * add visualize frame, visualize frames to codf w/ tests * add sample codf for tests * Format code with black and isort (#51) This commit fixes the style issues introduced in 87e2a98 according to the output from black and isort. Details: https://deepsource.io/gh/AtomScott/SoccerTrack/transform/59908cc6-7c0d-42cd-b45e-4b12743fac71/ Co-authored-by: deepsource-autofix[bot] <62050782+deepsource-autofix[bot]@users.noreply.github.com> * Format code with black and isort (#52) This commit fixes the style issues introduced in 87e2a98 according to the output from black and isort. Details: https://deepsource.io/gh/AtomScott/SoccerTrack/transform/d523db55-0b40-445f-b386-c3aa4e528174/ Co-authored-by: deepsource-autofix[bot] <62050782+deepsource-autofix[bot]@users.noreply.github.com> * Format code with black and isort (#53) This commit fixes the style issues introduced in dc69ea9 according to the output from black and isort. Details: https://deepsource.io/gh/AtomScott/SoccerTrack/transform/d454df1f-f632-42ef-b11e-6ed659e28f91/ Co-authored-by: deepsource-autofix[bot] <62050782+deepsource-autofix[bot]@users.noreply.github.com> * add documentation * add examples to documentation * fix: missing new line in documentation * Create CODE_OF_CONDUCT.md (#54) * Create publish-to-test-pypi.yml * update [tool.poetry] * Update pyproject.toml * Update pyproject.toml * Update pyproject.toml * testing changes Co-authored-by: Ikuma Uchida <[email protected]> Co-authored-by: IkumaUchida <[email protected]> Co-authored-by: DeepSource Bot <[email protected]> Co-authored-by: deepsource-autofix[bot] <62050782+deepsource-autofix[bot]@users.noreply.github.com>

IkumaUchida added 5 commits September 16, 2022 14:26

Added the function to visualize the bbox in one frame.

c591bd7

Added video-enabled visualization function

06906ae

Add Visualization Functions

d4375b7

Apply prospector to bboxdataframe

8339730

fix visualize_bbox function

329ba5e

AtomScott assigned IkumaUchida Sep 18, 2022

AtomScott commented Sep 18, 2022

View reviewed changes

IkumaUchida added 2 commits September 23, 2022 16:51

Merge branch 'feature/major_refactor' of github.com:AtomScott/SoccerT…

b3a0bcf

…rack into feature/visualization

fix visualization

0a33250

IkumaUchida reviewed Sep 25, 2022

View reviewed changes

AtomScott commented Sep 26, 2022

View reviewed changes

IkumaUchida added 12 commits September 27, 2022 09:29

fix visualization functions

9a30357

del index sort of visualize_frame in bboxdataframe

e23a15c

remove utils.py from PR

15003eb

fix utils and logging

a5b9bb0

fix utils

a89820f

fix utils

8f1757e

fix utils

87226d9

fix utils

0081fb8

fix logging

2b8f347

fix logging

3ff694c

fix logging

2d86a95

fix logging

642baba

AtomScott and others added 3 commits September 27, 2022 11:39

add comments to code

cc7547c

delete comments

970a004

fix visualization functions

e4a85fe

AtomScott merged commit 5842441 into feature/major_refactor Sep 27, 2022

AtomScott deleted the feature/visualization branch September 27, 2022 06:32

Feature/visualization （visualize_bbox functionに修正が必要なので） #32

Feature/visualization （visualize_bbox functionに修正が必要なので） #32

Conversation

AtomScott commented Sep 18, 2022

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

AtomScott commented Sep 18, 2022

review-notebook-app bot commented Sep 25, 2022

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

IkumaUchida commented Sep 25, 2022

AtomScott left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

IkumaUchida Sep 27, 2022 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

AtomScott Sep 26, 2022 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

AtomScott left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

AtomScott Sep 27, 2022 • edited Loading

Choose a reason for hiding this comment

AtomScott commented Sep 27, 2022

Feature/visualization　（visualize_bbox functionに修正が必要なので） #32

Feature/visualization　（visualize_bbox functionに修正が必要なので） #32

IkumaUchida Sep 27, 2022 •

edited

Loading

AtomScott Sep 26, 2022 •

edited

Loading

AtomScott Sep 27, 2022 •

edited

Loading