include action recognition and multi instance retrieval

epic_kitchens_100.py

@@ -14,7 +14,11 @@
 # limitations under the License.
 
 # Lint as: python3
-""""""
+"""EPIC-KITCHENS-100 is a large-scale dataset in first-person (egocentric) vision; multi-faceted, audio-visual,
+non-scripted recordings in native environments - i.e. the wearers' homes, capturing all daily activities
+in the kitchen over multiple days. Annotations are collected using a novel 'Pause-and-Talk' narration interface.
+
+EPIC-KITCHENS-100 is an extension of the EPIC-KITCHENS dataset released in 2018, to 100 hours of footage."""
 
 import os
 import csv
@@ -49,11 +53,22 @@ in the kitchen over multiple days. Annotations are collected using a novel 'Paus
 EPIC-KITCHENS-100 is an extension of the EPIC-KITCHENS dataset released in 2018, to 100 hours of footage.
 """
 
+_HOMEPAGE = "https://epic-kitchens.github.io/2022"
+
+_LICENSE = "CC BY-NC 4.0"
+
 _URL_BASE = "https://raw.githubusercontent.com/epic-kitchens/epic-kitchens-100-annotations/master/"
 
 
+_VARIANTS = [
+    "action_recognition", # This split is used by four challenges: Action Recognition, Weakly supervised action recognition, Action detection, Action anticipation
+    "multi_instance_retrieval",
+]
 class EpicKitchens100(datasets.GeneratorBasedBuilder):
-    """"""
+    """Epic Kitchens"""
+
+    BUILDER_CONFIGS = [datasets.BuilderConfig(name) for name in _VARIANTS]
+    DEFAULT_CONFIG_NAME = "action_recognition"
 
     def _info(self):
         return datasets.DatasetInfo(
@@ -64,117 +79,105 @@ class EpicKitchens100(datasets.GeneratorBasedBuilder):
                     "narration_id": datasets.Value("string"),
                     "participant_id": datasets.Value("string"),
                     "video_id": datasets.Value("string"),
+                    "path": datasets.Value("string"),
                     "narration_timestamp": datasets.Value("string"),
                     "start_timestamp": datasets.Value("string"),
                     "stop_timestamp": datasets.Value("string"),
-                    # "start_frame": datasets.Value("int32"),
-                    # "stop_frame": datasets.Value("int32"),
                     "narration": datasets.Value("string"),
                     "verb": datasets.Value("string"),
-                    "verb_class": datasets.Value("int32"), # More info: https://github.com/epic-kitchens/epic-kitchens-100-annotations/blob/master/README.md#epic_100_noun_classescsv
+                    "verb_class": datasets.Value("int32"),
+                    # The mapping for `verb_class` is available at: https://github.com/epic-kitchens/epic-kitchens-100-annotations/blob/master/README.md#epic_100_noun_classescsv
                     "noun": datasets.Value("string"),
-                    "noun_class": datasets.Value("string"), # More info: https://github.com/epic-kitchens/epic-kitchens-100-annotations/blob/master/README.md#epic_100_noun_classescsv
+                    "noun_class": datasets.Value("string"),
+                    # The mapping for `noun_class` is available at: https://github.com/epic-kitchens/epic-kitchens-100-annotations/blob/master/README.md#epic_100_noun_classescsv
                     "all_nouns": datasets.features.Sequence(datasets.Value("string")),
                     "all_noun_classes": datasets.features.Sequence(datasets.Value("int32")),
                 }
             ),
             supervised_keys=None,
-            homepage="https://epic-kitchens.github.io/2022",
+            homepage=_HOMEPAGE,
             citation=_CITATION,
+            license=_LICENSE
         )
 
     def _split_generators(self, dl_manager):
         urls = {
-            "annotations": {
+            "action_recognition": {
                 "train": os.path.join(_URL_BASE, "EPIC_100_train.csv"),
                 "validation": os.path.join(_URL_BASE, "EPIC_100_validation.csv"),
                 "test": os.path.join(_URL_BASE, "EPIC_100_test_timestamps.csv"),
             },
-            # "verb_and_noun_classes": {
-            #     "verb": os.path.join(_URL_BASE, "EPIC_100_verb_classes.csv"),
-            #     "noun": os.path.join(_URL_BASE, "EPIC_100_noun_classes.csv")
-            # }
+            "multi_instance_retrieval": {
+                "train": os.path.join(_URL_BASE, "retrieval_annotations/EPIC_100_retrieval_train.csv"),
+                "test": os.path.join(_URL_BASE, "retrieval_annotations/EPIC_100_retrieval_test.csv")
+            }
         }
         files_path = dl_manager.download_and_extract(urls)
-        return [
+        splits = [
             datasets.SplitGenerator(
                 name=datasets.Split.TRAIN,
                 gen_kwargs={
-                    "annotations": files_path["annotations"]["train"],
+                    "annotations": files_path[self.config.name]["train"],
                     "split": "train",
                 },
             ),
-            datasets.SplitGenerator(
-                name=datasets.Split.VALIDATION,
-                gen_kwargs={
-                    "annotations": files_path["annotations"]["validation"],
-                    "split": "validation",
-                },
-            ),
             datasets.SplitGenerator(
                 name=datasets.Split.TEST,
                 gen_kwargs={
-                    "annotations": files_path["annotations"]["test"],
+                    "annotations": files_path[self.config.name]["test"],
                     "split": "test",
                 },
             ),
         ]
-
-    # def _get_instances_mapping(self, file):
-    #     mapping = {}
-    #     with open(file, encoding="utf-8") as f:
-    #         csv_reader = csv.reader(f, delimiter=",")
-    #         next(csv_reader)
-    #         for id, key, instances, category in csv_reader:
-    #             mapping[int(id)] = {
-    #                 "key": key,
-    #                 "instances": instances,
-    #                 "category": category,
-    #             }
-    #     return mapping
+        if self.config.name == "action_recognition":
+            splits.append(
+                datasets.SplitGenerator(
+                    name=datasets.Split.VALIDATION,
+                    gen_kwargs={
+                        "annotations": files_path[self.config.name]["validation"],
+                        "split": "validation",
+                    },
+                ),
+            )
+        return splits
 
     def _generate_examples(self, annotations, split):
         """This function returns the examples."""
         with open(annotations, encoding="utf-8") as csv_file:
             csv_reader = csv.reader(csv_file, delimiter=",")
             next(csv_reader) # Skip header
-            if split != "test":
-                for idx, row in enumerate(csv_reader):
-                    yield idx, {
-                        "extended": len(row[0].split("_")[1]) == 3,
-                        "narration_id": row[0],
-                        "participant_id": row[1],
-                        "video_id": row[2],
-                        "narration_timestamp": row[3],
-                        "start_timestamp": row[4],
-                        "stop_timestamp": row[5],
-                        # "start_frame": row[6],
-                        # "stop_frame": row[7],
-                        "narration": row[8],
-                        "verb": row[9],
-                        "verb_class": row[10],
-                        "noun": row[11],
-                        "noun_class": row[12],
-                        "all_nouns": eval(row[13]),
-                        "all_noun_classes": eval(row[14]),
-                    }
-            else:
-                for idx, row in enumerate(csv_reader):
-                    yield idx, {
-                        "extended": len(row[0].split("_")[1]) == 3,
-                        "narration_id": row[0],
-                        "participant_id": row[1],
-                        "video_id": row[2],
-                        "narration_timestamp": row[3],
-                        "start_timestamp": row[4],
-                        "stop_timestamp": row[5],
-                        # "start_frame": row[6],
-                        # "stop_frame": row[7],
-                        "narration": "",
-                        "verb": "",
-                        "verb_class": -1,
-                        "noun": "",
-                        "noun_class": -1,
-                        "all_nouns": [],
-                        "all_noun_classes": [],
-                    }
+            for idx, row in enumerate(csv_reader):
+                narration_id, participant_id, video_id, narration_timestamp, start_timestamp, stop_timestamp = row[:6]
+                if split != "test":
+                    # The reason why it's jumping from 5 to 8 is that we are skipping `start_frame` and `stop_frame`
+                    # since we are not exposing the frames, but just the videos
+                    narration, verb, verb_class, noun, noun_class, all_nouns, all_noun_classes = row[8:15]
+                    all_nouns = eval(all_nouns)
+                    all_noun_classes = eval(all_noun_classes)
+                else:
+                    narration = verb = noun = ""
+                    verb_class = noun_class = -1
+                    all_nouns = all_noun_classes = []
+                extended = len(narration_id.split("_")[1]) == 3
+                if extended:
+                    path = f"EPIC-KITCHENS/{participant_id}/videos/{video_id}.MP4" #Paths for jz version
+                else:
+                    path = f"EPIC_KITCHENS_2018/videos/{split}/{participant_id}/{video_id}.MP4" #Paths for jz version
+
+                yield idx, {
+                    "extended": extended,
+                    "narration_id": narration_id,
+                    "participant_id": participant_id,
+                    "video_id": video_id,
+                    "path": path,
+                    "narration_timestamp": narration_timestamp,
+                    "start_timestamp": start_timestamp,
+                    "stop_timestamp": stop_timestamp,
+                    "narration": narration,
+                    "verb": verb,
+                    "verb_class": verb_class,
+                    "noun": noun,
+                    "noun_class": noun_class,
+                    "all_nouns": all_nouns,
+                    "all_noun_classes": all_noun_classes,
+                }