Cross-Modal Information Retrieval (CMIR) is the task of finding relevant items across different modalities. For example, given an image, find a text or vice versa. The main challenge in CMIR is known as the heterogeneity gap: since items from different modalities have different data types, the similarity between them cannot be measured directly. Therefore, the majority of CMIR methods published to date attempt to bridge this gap by learning a latent representation space, where the similarity between items from different modalities can be measured.
The TED VCR Video Retrieval Dataset is a multimodal collection derived from publicly available TED Talks. It contains thousands of talks filtered to retain only those with meaningful topic labels, producing a long-tail, multi-label taxonomy. For each talk the dataset provides automatic speech-recognition transcripts, slide- and scene-level OCR text, and frame-level visual captions—three textual channels used in VCR retrieval experiments. The data are split into 80 % train, 10 % validation, and 10 % test while preserving the original topic distribution, leaving 542 talks as a held-out test set. Two ready-to-download archives accompany the release: 4.2 GB of trimmed MP4 videos with metadata and 1.8 GB of pre-computed CLIP and Whisper embeddings, both shared under the non-commercial CC BY-NC-ND 4.0 license.