Babel-ImageNet: Massively Multilingual Evaluation of Vision-and-Language Representations

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Geigle, Gregor, Timofte, Radu, Glavaš, Goran
Format: Preprint
Published: 2023
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866910482513788928
author Geigle, Gregor
Timofte, Radu
Glavaš, Goran
author_facet Geigle, Gregor
Timofte, Radu
Glavaš, Goran
contents Vision-and-language (VL) models with separate encoders for each modality (e.g., CLIP) have become the go-to models for zero-shot image classification and image-text retrieval. They are, however, mostly evaluated in English as multilingual benchmarks are limited in availability. We introduce Babel-ImageNet, a massively multilingual benchmark that offers (partial) translations of ImageNet labels to 100 languages, built without machine translation or manual annotation. We instead automatically obtain reliable translations by linking them -- via shared WordNet synsets -- to BabelNet, a massively multilingual lexico-semantic network. We evaluate 11 public multilingual CLIP models on zero-shot image classification (ZS-IC) on our benchmark, demonstrating a significant gap between English ImageNet performance and that of high-resource languages (e.g., German or Chinese), and an even bigger gap for low-resource languages (e.g., Sinhala or Lao). Crucially, we show that the models' ZS-IC performance highly correlates with their performance in image-text retrieval, validating the use of Babel-ImageNet to evaluate multilingual models for the vast majority of languages without gold image-text data. Finally, we show that the performance of multilingual CLIP can be drastically improved for low-resource languages with parameter-efficient language-specific training. We make our code and data publicly available: \url{https://github.com/gregor-ge/Babel-ImageNet}
format Preprint
id arxiv_https___arxiv_org_abs_2306_08658
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle Babel-ImageNet: Massively Multilingual Evaluation of Vision-and-Language Representations
Geigle, Gregor
Timofte, Radu
Glavaš, Goran
Computation and Language
Computer Vision and Pattern Recognition
Vision-and-language (VL) models with separate encoders for each modality (e.g., CLIP) have become the go-to models for zero-shot image classification and image-text retrieval. They are, however, mostly evaluated in English as multilingual benchmarks are limited in availability. We introduce Babel-ImageNet, a massively multilingual benchmark that offers (partial) translations of ImageNet labels to 100 languages, built without machine translation or manual annotation. We instead automatically obtain reliable translations by linking them -- via shared WordNet synsets -- to BabelNet, a massively multilingual lexico-semantic network. We evaluate 11 public multilingual CLIP models on zero-shot image classification (ZS-IC) on our benchmark, demonstrating a significant gap between English ImageNet performance and that of high-resource languages (e.g., German or Chinese), and an even bigger gap for low-resource languages (e.g., Sinhala or Lao). Crucially, we show that the models' ZS-IC performance highly correlates with their performance in image-text retrieval, validating the use of Babel-ImageNet to evaluate multilingual models for the vast majority of languages without gold image-text data. Finally, we show that the performance of multilingual CLIP can be drastically improved for low-resource languages with parameter-efficient language-specific training. We make our code and data publicly available: \url{https://github.com/gregor-ge/Babel-ImageNet}
title Babel-ImageNet: Massively Multilingual Evaluation of Vision-and-Language Representations
topic Computation and Language
Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2306.08658