Add files using upload-large-folder tool

README.md

@@ -1,6 +1,6 @@
-# SO-Dataset: Spatial FOA Audio with FSD63 Event Labels
+# SO-Dataset: Spatial FOA Audio Dataset
 
-SO-Dataset is a large-scale spatial audio dataset in first-order ambisonics (FOA) format. Each example contains one FOA waveform and spatial event annotations in DCASE-style CSV files. The dataset combines simulated spatial scenes and real FOA recordings, and all sound event labels are mapped into a unified 63-class FSD taxonomy.
+SO-Dataset is a large-scale spatial audio dataset in first-order ambisonics (FOA) format. Each example contains FOA waveform and spatial event annotations in DCASE-style CSV files. The dataset combines simulated spatial scenes and real FOA recordings, and all sound event labels are mapped into a unified 63-class sound event taxonomy based on the FSD50k dataset.
 
 The public release stores audio and annotations as tar shards. The tar files preserve the same relative paths used by the metadata, so extracting the archives recreates the `audio/` and `annotations/` directories expected by the JSONL files.
 
@@ -8,7 +8,7 @@ The public release stores audio and annotations as tar shards. The tar files pre
 
 - **Audio format**: FOA waveform files (`.wav`)
 - **Spatial annotations**: DCASE-style CSV files
-- **Labels**: unified FSD63 label set
+- **Labels**: unified FSD50k label set
 - **Splits**: `train`, `valid`, `test`
 - **Metadata**: one JSON object per audio scene
 - **Packaging**: path-preserving tar shards for easier download and upload
@@ -61,25 +61,13 @@ These paths match the paths stored in `metadata/*.jsonl`.
 
 ## Dataset Statistics
 
-| Split | Scenes / FOA files | Audio shards | Annotation files | Annotation shards |
-|---|---:|---:|---:|---:|
-| train | 329,610 | 188 | 1,032,417 | 1 |
-| valid | 35,093 | 23 | 127,028 | 1 |
-| test | 35,237 | 22 | 106,460 | 1 |
-| total | 399,940 | 233 | 1,265,905 | 3 |
+SO-Dataset contains 400K FOA audio segments across 233 scenes, with a total of 1.27M annotated sound events.
 
-Total tar files: **236**.
+The dataset and annotations details are shown in the following figure.
+Figure(a) shows the sub-tasks in SO-QA and SO-Bench, including Detection and Localization, Spatial Relation Understanding, and Complex Reasoning with Semantics. Figure(b) shows the data source of sound events in the dataset. Figure(c) shows the building process of the dataset, including the recording, simulation and collect subset. After building the SO-Dataset, we generate QA pairs and build SO-QA using the metadata of SO-Dataset. Figure(d) shows the distribution of spatial event in our dataset, including azimuth, elevation and distance. 
 
-Approximate payload sizes:
+![Dataset statistics](figures/data.png)
 
-| Group | Payload bytes |
-|---|---:|
-| audio/train | 939,040,161,296 |
-| audio/valid | 110,868,182,152 |
-| audio/test | 108,399,164,604 |
-| annotations/train | 2,350,160,868 |
-| annotations/valid | 313,645,376 |
-| annotations/test | 263,433,222 |
 
 ## Metadata Format
 
@@ -134,17 +122,18 @@ Important fields:
 
 ## Label Mapping
 
-`label_mapping.json` defines the FSD63 label space:
+`label_mapping.json` defines the sound event label space:
 
-- `class_set`: `fsd63`
 - `class_count`: `63`
 - `class_names`: ordered class names
 - `class_name_to_id`: class name to integer id
 - `class_id_to_name`: integer id to class name
-- `raw_label_aliases`: aliases used before final mapping, such as mapping singing variants to `singing`
 
 Use `label_mapping.json` as the canonical class-id definition.
 
+The following figure shows the distribution of the 63 classes in our dataset.  
+![Class distribution](figures/event.png)
+
 ## Download
 
 Install the Hugging Face CLI:
@@ -212,29 +201,6 @@ annotations/train/foa_....csv
 annotations/train/foa_...._src00.csv
 ```
 
-## Reading Metadata
-
-```python
-import json
-from pathlib import Path
-
-root = Path("SO-Dataset")
-
-with (root / "metadata" / "train.jsonl").open("r", encoding="utf-8") as f:
-    item = json.loads(next(f))
-
-audio_path = root / item["audio"]["foa_path"]
-scene_csv_path = root / item["scene_annotation_csv_path"]
-source_csv_paths = [
-    root / src["source_trajectory_csv_path"]
-    for src in item["sources"]
-]
-
-print(audio_path)
-print(scene_csv_path)
-print(source_csv_paths[:3])
-```
-
 ## Manifests
 
 The files in `manifests/` list the tar shards for each group and split.
@@ -252,14 +218,19 @@ Example row:
 }
 ```
 
-## Notes
-
-- The release is anonymized: public audio and CSV filenames use hashed names.
-- Metadata paths are relative to the dataset root.
-- The tar archives are not compressed. This keeps extraction fast and avoids heavy CPU cost for already-large waveform data.
-- Scene-level CSV files contain the combined annotation for all active sources in a FOA scene.
-- Per-source CSV files are also provided and are referenced from `sources[*].source_trajectory_csv_path`.
 
 ## Citation and License
 
 Please cite this dataset as appropriate for your use. If you redistribute or use the dataset in downstream work, make sure your usage is compatible with the licenses of the underlying audio and spatial data sources.
+
+```text
+@misc{zhu2026spatialomnispatialaudiounderstanding,
+      title={Spatial-Omni: Spatial Audio Understanding Integration in Multimodal LLMs via FOA Encoding}, 
+      author={Zhiyuan Zhu and Yixuan Chen and Yiwen Shao and Wenxiang Guo and Changhao Pan and Yu Zhang and Yuxiang Wang and Wei Liu and Houhua Zhang and Chengkuan Zeng and Wenbo Cheng and Yunxi Liu and Rui Yang and Steve Yves and Liefeng Bo and Zhou Zhao},
+      year={2026},
+      eprint={2606.10738},
+      archivePrefix={arXiv},
+      primaryClass={eess.AS},
+      url={https://arxiv.org/abs/2606.10738}, 
+}
+```