SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2406.11546 · 2024

GigaSpeech 2: An Evolving, Large-Scale and Multi-domain ASR Corpus for Low-Resource Languages with Automated Crawling, Transcription and Refinement

arXiv · PDF · Open in the Atlas

Code that ran

1 of 2 samples (functions and classes) lifted from this paper's linked repositories executed on a synthesized input in Syntology's jail (0 of them constructed an object rather than computing a result; 1 honoured the contract we drafted, 0 violated it and 0 have no contract verdict; 0 failures of our instrument; 1 no recorded run), measured 2026-11-22; repository links from the Papers with Code archive attributed under CC BY-SA, no affiliation; bodies inline only under a permissive licence as GitHub detects it. Not a reproduction of the paper's results.

"Ran" means the function executed on a synthesized input and returned a value. It is not a reproduction of the paper's results. This record, dated, as JSON: /api/ran/2406.11546.json · badge: /api/ran/2406.11546.svg

Each repository's role is a label: a classifier's output recorded on the link, or, where no role is recorded, what the Papers with Code archive marked. The line under it names its basis. Neither says who wrote the code.

RepositoryRoleRan
SpeechColab/GigaSpeech2 canonical (classifier label)
classifier basis: classifier_2025
1 ran · 1 no recorded run
FunctionStatusWhere it lives
read_wav_info Ran
ran, honoured the contract we drafted
SpeechColab/GigaSpeech2/pipeline/convert_transcribe/convert_and_transcribe.py
code served (permissive licence) · get_code("fc3617f5760e01d7")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1) and returned tuple, ndim 1, dtype family float; no output shape is stated. The image it ran in was not recorded. If your copy returns the same type and ndim on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import struct
import tempfile
import os

# Create a minimal valid WAV file for testing
def create_test_wav():
    """Create a minimal WAV file with proper headers"""
    temp_file = tempfile.NamedTemporaryFile(delete=False, suffix='.wav')
    filename = temp_file.name
    
    # WAV file structure:
    # 0-3: "RIFF"
    # 4-7: file size - 8 (little-endian int)
    # 8-11: "WAVE"
    # 12-15: "fmt "
    # 16-19: subchunk1 size (16 for PCM)
    # 20-21: audio format (1 for PCM)
    # 22-23: number of channels (little-endian short)
    # 24-27: sample rate (little-endian int)
    # 28-31: byte rate
    # 32-33: block align
    # 34-35: bits per sample
    # 36-39: "data"
    # 40-43: data size
    
    sr = 16000  # sample rate
    channels = 2  # stereo
    bits_per_sample = 16
    duration_seconds = 1
    
    num_samples = sr * duration_seconds
    data_size = num_samples * channels * (bits_per_sample // 8)
    
    # Build WAV header
    wav_data = bytearray()
    
    # RIFF header
    wav_data.extend(b'RIFF')
    file_size = 36 + data_size
    wav_data.extend(struct.pack('<I', file_size))
    wav_data.extend(b'WAVE')
    
    # fmt subchunk
    wav_data.extend(b'fmt ')
    wav_data.extend(struct.pack('<I', 16))  # subchunk1 size
    wav_data.extend(struct.pack('<H', 1))   # audio format (PCM)
    wav_data.extend(struct.pack('<H', channels))  # number of channels
    wav_data.extend(struct.pack('<I', sr))  # sample rate
    byte_rate = sr * channels * (bits_per_sample // 8)
    wav_data.extend(struct.pack('<I', byte_rate))  # byte rate
    block_align = channels * (bits_per_sample // 8)
    wav_data.extend(struct.pack('<H', block_align))  # block align
    wav_data.extend(struct.pack('<H', bits_per_sample))  # bits per sample
    
    # data subchunk
    wav_data.extend(b'data')
    wav_data.extend(struct.pack('<I', data_size))  # data size
    
    # Add dummy audio data
    wav_data.extend(b'\x00' * data_size)
    
    temp_file.write(wav_data)
    temp_file.close()
    
    return filename

wav_filename = create_test_wav()

ARGS = {
    'filename': wav_filename
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1
Imports
none outside the standard library — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type tuple, ndim 1, dtype family float — a tuple output has no array shape of its own; none is stated
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
wav2whisper Not yet run
unverified (harvested, no recorded run)
SpeechColab/GigaSpeech2/pipeline/convert_transcribe/convert_and_transcribe.py
code served (permissive licence) · get_code("11fa222060b15a90")
How it ran

no invocation recorded.

Also present in other papers' code

Of 2 samples lifted from this paper's linked repositories, 2 carry a function identity and 0 do not. Of the 2 distinct functions with an identity, 0 are also present in other papers whose code we hold (0 with at least one paper dated earlier, 0 the same month, 0 later) and 2 are only here among the 52,977 papers whose code Syntology holds (as of 2026-09-26).

Only here means: among the 52,977 papers whose code Syntology holds (51,887 with lifted samples, 1,090 through their assemblies only), as of 2026-09-26, no other paper's lifted samples or current assembly bricks contain an identical function body after normalisation.

Presence, not direction: an identical function body in two papers' repositories does not say who wrote it, which came from which, or that either paper's method rests on it. Matching is exact after normalisation, so a renamed or edited function is not matched. Dates are when papers appeared, not when code was written. Only .py files the harvester lifted are covered; notebooks are not read. A new-style arXiv id YYMM.NNNN dates a paper to the month 20YY-MM; otherwise Paper.year dates it to the year; otherwise it is undated. Dates are when the paper appeared, not when the code was written. Read 2026-11-22.

Repositories linked to this paper

Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.

Abstract

The evolution of speech technology has been spurred by the rapid increase in dataset sizes. Traditional speech models generally depend on a large amount of labeled training data, which is scarce for low-resource languages. This paper presents GigaSpeech 2, a large-scale, multi-domain, multilingual speech recognition corpus. It is designed for low-resource languages and does not rely on paired speech and text data. GigaSpeech 2 comprises about 30,000 hours of automatically transcribed speech, including Thai, Indonesian, and Vietnamese, gathered from unlabeled YouTube videos. We also introduce an automated pipeline for data crawling, transcription, and label refinement. Specifically, this pipeline involves Whisper for initial transcription, MMS for forced alignment, and multi-dimensional filtering for data quality assurance. A modified Noisy Student Training is developed to further refine flawed pseudo labels iteratively, thereby enhancing model performance. Experimental results on our manually transcribed evaluation set and two public test sets from Common Voice and FLEURS confirm our corpus's high quality and broad applicability. Notably, ASR models trained on GigaSpeech 2 can reduce the word error rate for Thai, Indonesian, and Vietnamese on our challenging and realistic YouTube test set by 25% to 40% compared to Whisper large-v3, with merely 10% model parameters. Furthermore, our ASR models trained on GigaSpeech 2 yield superior performance compared to commercial services. We hope that our newly introduced corpus and pipeline will open a new avenue for low-resource speech recognition and significantly facilitate research in this area.

For agents

The same record, over MCP at https://syntology.ai/mcp:

get_harvested_code_for_paper("2406.11546")
get_code_for_paper("2406.11546")
have("2406.11546")

The run record, dated, one paper per request, free:

curl https://syntology.ai/api/ran/2406.11546.json

A badge for a README (the split and the date, never a ratio):

[![Syntology run record](https://syntology.ai/api/ran/2406.11546.svg)](https://syntology.ai/paper/2406.11546)

Connect an agent — have() is free.