Dataset Viewer

The dataset viewer is not available for this subset.

Cannot get the split names for the config 'default' of the dataset.

Exception:    SplitsNotFoundError
Message:      The split names could not be parsed from the dataset config.
Traceback:    Traceback (most recent call last):
                File "/usr/local/lib/python3.14/site-packages/datasets/inspect.py", line 286, in get_dataset_config_info
                  for split_generator in builder._split_generators(
                                         ~~~~~~~~~~~~~~~~~~~~~~~~~^
                      StreamingDownloadManager(base_path=builder.base_path, download_config=download_config)
                      ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
                  )
                  ^
                File "/usr/local/lib/python3.14/site-packages/datasets/packaged_modules/folder_based_builder/folder_based_builder.py", line 246, in _split_generators
                  raise ValueError(
                      "`file_name`, `*_file_name`, `file_names` or `*_file_names` must be present as dictionary key in metadata files"
                  )
              ValueError: `file_name`, `*_file_name`, `file_names` or `*_file_names` must be present as dictionary key in metadata files
              
              The above exception was the direct cause of the following exception:
              
              Traceback (most recent call last):
                File "/src/services/worker/src/worker/job_runners/config/split_names.py", line 68, in compute_split_names_from_streaming_response
                  for split in get_dataset_split_names(
                               ~~~~~~~~~~~~~~~~~~~~~~~^
                      path=dataset,
                      ^^^^^^^^^^^^^
                      config_name=config,
                      ^^^^^^^^^^^^^^^^^^^
                      token=hf_token,
                      ^^^^^^^^^^^^^^^
                  )
                  ^
                File "/usr/local/lib/python3.14/site-packages/datasets/inspect.py", line 340, in get_dataset_split_names
                  info = get_dataset_config_info(
                      path,
                  ...<6 lines>...
                      **config_kwargs,
                  )
                File "/usr/local/lib/python3.14/site-packages/datasets/inspect.py", line 291, in get_dataset_config_info
                  raise SplitsNotFoundError("The split names could not be parsed from the dataset config.") from err
              datasets.inspect.SplitsNotFoundError: The split names could not be parsed from the dataset config.

Need help to make the dataset viewer work? Make sure to review how to configure the dataset viewer, and open a discussion for direct support.

Dremel Cut Quality Dataset

A multimodal dataset of rotary-tool (Dremel) cutting and sanding operations on metal workpieces, for surface-defect detection. Every example pairs a photograph of the finished cut with the audio recorded during the operation, and carries labels at three levels: a clip-level normal/abnormal verdict, a per-window audio label, and a spatial grid label on the image.

Summary


Examples	3091
Clip label	1881 normal, 1210 abnormal
Fault types (abnormal)	dent 574, wave 348, scuff 147, partial_scuff 141
Toolhead	cutting_disc 2069, sanding_cut 1022
Audio	48 kHz, mono, 16-bit PCM; 0.5–8.0 s (median 1.5 s)
Image	RGB JPEG, width 102–1465 px, height 58–1031 px

Layout

metadata.csv
example_0001/
    image.jpg
    audio.wav
    audio_sam_target.wav
    audio_label.npy
    image_label.npy
example_0002/
    ...

Files

image.jpg — photograph of the workpiece after the operation, cropped to the cut region. Aspect ratio varies with cut length.

audio.wav — the sound recorded during the operation. Single channel, 48 kHz, 16-bit PCM.

audio_sam_target.wav — a source-separated version of audio.wav that isolates the tool sound from background noise, produced with SAM-Audio. Same format, length, and timebase as audio.wav.

audio_label.npy — int8 array of shape [T], one entry per consecutive 0.5 s window of audio.wav (T = ceil(duration / 0.5), the final window may be shorter). Values: 0 normal, 1 defect.

image_label.npy — int8 array of shape [16, 16], a grid laid over image.jpg in row-major order (cell = row * 16 + col). Values: 1 defect, 0 clean surface, -1 ignore (outside the workpiece or not annotated). Normal examples are all 0.

metadata.csv

column	description
`example_id`	directory name, `example_0001` … `example_3091`
`label`	`normal` or `abnormal`
`fault_type`	`dent`, `wave`, `scuff`, `partial_scuff`; empty for normal
`toolhead`	`cutting_disc` or `sanding_cut`
`duration_s`	length of `audio.wav` in seconds

Loading

import csv, numpy as np, soundfile as sf
from pathlib import Path

root = Path("Dremel_Final_Dataset")
meta = {r["example_id"]: r for r in csv.DictReader(open(root / "metadata.csv"))}

ex = root / "example_0001"
audio, sr = sf.read(ex / "audio.wav")          # (N,) float64, sr = 48000
y_audio = np.load(ex / "audio_label.npy")      # (T,)  int8
y_image = np.load(ex / "image_label.npy")      # (16, 16) int8
clip_label = meta["example_0001"]["label"]

Label notes

About 21% of audio windows in the dataset are defect; the rest are normal. Normal clips contain only normal windows.
In image_label.npy, treat -1 as an ignore index for spatial losses and metrics. Abnormal examples always contain at least one 1 cell.
Every normal clip has an all-0 audio_label.npy and image_label.npy. Every abnormal clip has at least one 1 cell in image_label.npy and at least one 1 window in audio_label.npy.
fault_type is a property of the whole clip, not of individual cells or windows.

Downloads last month: 428

Total file size:

966 MB