Upload folder using huggingface_hub

README.md

@@ -0,0 +1,116 @@
+---
+configs:
+  - config_name: default
+    data_files:
+      - split: test
+        path: E2E_VP_default_test.json
+license: mit
+task_categories:
+  - video-classification
+  - question-answering
+language:
+  - en
+tags:
+  - video
+  - spatial-reasoning
+  - direction
+  - synthetic
+  - VideoLLM
+pretty_name: E2E 20K Synthetic Direction
+size_categories:
+  - 10K<n<100K
+---
+
+# E2E 20K Synthetic Direction
+
+A large-scale synthetic video benchmark for evaluating VideoLLMs' directional reasoning.  
+20,000 videos of **colored geometric shapes** moving in four cardinal directions, with automatically generated MCQ annotations.
+
+## Directions
+
+| Direction | Description |
+|---|---|
+| `up` | Object moves toward the top of the frame |
+| `down` | Object moves toward the bottom of the frame |
+| `left` | Object moves toward the left of the frame |
+| `right` | Object moves toward the right of the frame |
+
+## Usage
+
+```python
+from datasets import load_dataset
+
+ds = load_dataset("YOUR_HF_ID/E2E_20K", split="train")
+```
+
+## Data Format
+
+```json
+{
+  "id": 0,
+  "video": "synthetic20k/train/down/circle_blue_001.mp4",
+  "question": "Which direction does the blue circle move (camera perspective)?",
+  "candidates": ["Left", "Down", "Right", "Up"],
+  "answer": "B",
+  "answer_text": "Down",
+  "direction": "down"
+}
+```
+
+| Field | Type | Description |
+|---|---|---|
+| `id` | int | 고유 인덱스 |
+| `video` | str | 비디오 상대 경로 |
+| `question` | str | 방향 질문 (question template 랜덤 샘플링) |
+| `candidates` | list[str] | 4개 보기 (순서 랜덤 셔플) |
+| `answer` | str | 정답 보기 레이블 (`A`/`B`/`C`/`D`) |
+| `answer_text` | str | 정답 텍스트 (`Up` / `Down` / `Left` / `Right`) |
+| `direction` | str | GT 방향 소문자 |
+
+## Video Structure
+
+```
+synthetic20k/
+├── train/
+│   ├── up/
+│   │   ├── circle_blue_001.mp4
+│   │   ├── circle_red_001.mp4
+│   │   └── ...
+│   ├── down/
+│   ├── left/
+│   └── right/
+└── val/
+    ├── up/
+    ├── down/
+    ├── left/
+    └── right/
+```
+
+## Shapes & Colors
+
+**Shapes:** circle, square, triangle  
+**Colors:** blue, red, green, yellow, white
+
+## Statistics
+
+| Split | Total | Up | Down | Left | Right |
+|---|---|---|---|---|---|
+| train | 16,000 | 4,000 | 4,000 | 4,000 | 4,000 |
+| val | 4,000 | 1,000 | 1,000 | 1,000 | 1,000 |
+| **total** | **20,000** | **5,000** | **5,000** | **5,000** | **5,000** |
+
+## Question Templates
+
+Each sample draws from a pool of question templates to reduce template bias:
+
+- *"Which direction does the {color} {shape} move (camera perspective)?"*
+- *"As the clip progresses, which direction does the {color} {shape} move?"*
+- *"Which direction does the {color} {shape} move within the frame?"*
+- *"Which way does the {color} {shape} move in the image plane?"*
+- *(and more...)*
+
+## Related Datasets
+
+| Dataset | Description |
+|---|---|
+| [E2E_real_object](#) | Real-world objects, same direction task, smaller scale |

convert.py

@@ -0,0 +1,83 @@
+import json
+import re
+import argparse
+
+def parse_content(content: str):
+    """messages[0].content에서 question과 choices 파싱"""
+    # Question 추출
+    question_match = re.search(r'Question: (.+?)\n', content)
+    question = question_match.group(1).strip() if question_match else ""
+
+    # Choices 추출: "- The X is moving Y." 형태에서 Y만 뽑기
+    choices_raw = re.findall(r'- .+ is moving (\w+)\.', content)
+    candidates = [c.capitalize() for c in choices_raw]
+
+    return question, candidates
+
+import random
+
+def convert(input_path: str, output_path: str, video_prefix: str, ratio: float = 1.0):
+    with open(input_path, "r") as f:
+        data = json.load(f)
+
+    if not (0.0 < ratio <= 1.0):
+        raise ValueError(f"--ratio는 0.0 초과 1.0 이하여야 합니다. (입력값: {ratio})")
+
+    if ratio < 1.0:
+        k = max(1, int(len(data) * ratio))
+        data = random.sample(data, k)
+        print(f"🎲 {len(data)}개 샘플링 (ratio={ratio})")
+
+    results = []
+    for idx, item in enumerate(data):
+        content = item["messages"][0]["content"]
+        assistant_text = item["messages"][1]["content"]  # e.g. "The blue circle is moving down."
+        video_src = item["videos"][0]                    # e.g. "synthetic20k/train/down/circle_blue_001.mp4"
+
+        question, candidates = parse_content(content)
+
+        # answer_text: assistant 답변에서 방향 추출
+        dir_match = re.search(r'moving (\w+)\.', assistant_text)
+        answer_text = dir_match.group(1).capitalize() if dir_match else ""
+        direction = answer_text.lower()
+
+        # answer: candidates 내 위치 → A/B/C/D
+        try:
+            answer_idx = candidates.index(answer_text)
+            answer = chr(ord("A") + answer_idx)
+        except ValueError:
+            answer = "?"
+
+        # video 경로: 접두사 교체 (train/val 포함)
+        # e.g. synthetic20k/train/down/circle_blue_001.mp4
+        #   -> E2E_VP_default/val/down/circle_blue_001.mp4
+        parts = video_src.split("/")  # ['synthetic20k', 'train', 'down', 'circle_blue_001.mp4']
+        new_video = "/".join([video_prefix] + parts[2:])  # prefix + down/circle_blue_001.mp4
+
+        results.append({
+            "id": idx,
+            "video": new_video,
+            "question": question,
+            "candidates": candidates,
+            "answer": answer,
+            "answer_text": answer_text,
+            "direction": direction,
+        })
+
+    with open(output_path, "w") as f:
+        json.dump(results, f, indent=2, ensure_ascii=False)
+
+    print(f"✅ {len(results)}개 변환 완료 → {output_path}")
+
+
+if __name__ == "__main__":
+    parser = argparse.ArgumentParser()
+    parser.add_argument("--input",  required=True, help="원본 JSON 경로")
+    parser.add_argument("--output", required=True, help="출력 JSON 경로")
+    parser.add_argument("--prefix", default="E2E_VP_default/val",
+                        help="새 video 경로 접두사 (default: E2E_VP_default/val)")
+    parser.add_argument("--ratio", type=float, default=1.0,
+                        help="샘플링 비율 0.0~1.0 (default: 1.0 = 전체)")
+    args = parser.parse_args()
+
+    convert(args.input, args.output, args.prefix, args.ratio)