Cerru02/LombardGrid-I2VA

Lombard GRID Image-to-Video+Audio Speaker Retrieval

An MTEB/MOEB image-to-video+audio (i2va) speaker-retrieval benchmark derived from the Lombard GRID corpus. Query images show a frontal view of a talker. Corpus items contain a different utterance from the side/profile camera together with the matching separate high-quality audio recording. Every corpus item from the same speaker is relevant.

The source paper introduced the corpus but did not define this retrieval task. Relevance is derived from native speaker identity labels.

Frozen evaluation protocol

• Test queries: 540 (10 per speaker; 5 plain and 5 Lombard)
• Test corpus: 1080 (20 per speaker; 10 plain and 10 Lombard)
• Binary qrels: 10800
• Speakers: 54 (s2 through s55; s1 was excluded by the source)
• Relevant corpus items per query: 20
• Distinct utterance codes per speaker: 30
• Query/corpus utterance-code overlap: 0
• Selection seed: mteb-lombard-grid-i2va-v1

Within each speaker and condition, recordings are ordered by SHA-256 over the fixed seed, role, and source filename. The query and corpus selections are balanced across plain and Lombard conditions and are disjoint at the (speaker, utterance_code) level. Query images are the frame displayed at the temporal midpoint of the selected frontal clip. Opaque IDs are published; speaker, condition, utterance code, transcription, and other text are not dataset features.

Source audit and release discrepancies

The official JSON archive contains 5,340 rows, while each media archive contains 5,390 files. All three media archives have exactly the same recording filenames. The media restore 50 s31 recordings absent from JSON. s51 has only 40 plain and 50 Lombard files, rather than the advertised 50 of each. s14 has 51 plain and 49 Lombard rows, including a duplicate prwr3a recording key. The JSON metadata labels 69 recordings as WRONG. Media filenames mark 70: the additional one is among the 50 media-only s31 recordings. All 70, plus the one legacy-named duplicate, are excluded from selection. Four frontal MOV files lack a moov atom and are corrupt: s32_l_pwip9p, s32_p_bwwj2n, s33_l_pwajza, and s33_p_sgwq2s.

The deterministic selection required 0 fallback(s). The selected protocol remains exactly balanced after exclusions.

Media processing and validation

The released MOV files are already H.264/yuv420p visual-only clips, so 1080 selected corpus videos are copied without transcoding. 0 video(s) required a lossless video-stream remux to remove embedded audio. Separate source WAV files are copied without transcoding. Only the query midpoint frames are newly encoded, as RGB PNG files.

All 540 images, all 1080 complete videos, and all 1080 complete audio files were decoded during construction. Matching speaker, condition, utterance code, and source filename stems establish audiovisual correspondence. Selected video/audio duration differences were checked and paired decoding succeeded. The source paper reports correlation-based audiovisual alignment before utterance extraction; this construction did not independently verify signal-level synchronization.

Configs follow MTEB's retrieval layout:

• queries/test: id, image
• corpus/test: id, video, audio
• qrels/test: query-id, corpus-id, score

Evaluation

The primary metric is nDCG@10. Other standard MTEB retrieval metrics are reported as secondary results. This multi-positive speaker-retrieval protocol is newly derived, so there is no source-paper retrieval score to reproduce.

License and attribution

The source corpus is released under the Creative Commons Attribution 4.0 International license. Please attribute the original authors and cite the paper below. The MTEB task metadata and this card both use cc-by-4.0.

Responsible use

This task evaluates biometric speaker identity using faces and voices. Such representations can create privacy, surveillance, demographic-bias, and misidentification risks. Results should be treated as research measurements, not as evidence that a system is suitable for identity decisions. Users should consider participant consent, applicable biometric-data law, subgroup behavior, and downstream misuse before training or deploying related systems.

Citation

@article{alghamdi2018corpus,
  author = {Alghamdi, Najwa and Maddock, Steve and Marxer, Ricard and Barker, Jon and Brown, Guy J.},
  title = {A corpus of audio-visual Lombard speech with frontal and profile views},
  journal = {The Journal of the Acoustical Society of America},
  volume = {143},
  number = {6},
  pages = {EL523--EL529},
  year = {2018},
  doi = {10.1121/1.5042758},
}