We introduce GLAMI-1M: the largest multilingual image-text classification
dataset and benchmark. The dataset contains images of fashion products with
item descriptions, each in 1 of 13 languages. Categorization into 191 classes
has high-quality annotations: all 100k images in the test set and 75% of the 1M
training set were human-labeled. The paper presents baselines for image-text
classification showing that the dataset presents a challenging fine-grained
classification problem: The best scoring EmbraceNet model using both visual and
textual features achieves 69.7% accuracy. Experiments with a modified Imagen
model show the dataset is also suitable for image generation conditioned on
text. The dataset, source code and model checkpoints are published at
this https URL

GLAMI-1M 是一个包含多语言、时尚产品图像及描述的最大分类数据集，共包含 191 个类别，通过人工标注的测试集和部分训练集对其进行高质量注释。该数据集可用于图像生成和文本分类，最佳 EmbraceNet 模型在使用视觉和文本特征的情况下达到了 69.7% 的准确率并展示了其具有挑战性的精细分类问题。

GLAMI-1M: 多语言图像文本时尚数据集

GLAMI-1M: A Multilingual Image-Text Fashion Dataset

In this paper, we report an industry-academia collaborative study on the
distribution method of fashion products using an artificial intelligence (AI)
technique combined with an optimization method. To meet the current fashion
trend of short product lifetimes and an increasing variety of styles, the
company produces limited volumes of a large variety of styles. However, due to
the limited volume of each style, some styles may not be distributed to some
off-line stores. As a result, this high-variety, low-volume strategy presents
another challenge to distribution managers. We collaborated with KOLON F/C, one
of the largest fashion business units in South Korea, to develop models and an
algorithm to optimally distribute the products to the stores based on the
visual images of the products. The team developed a deep learning model that
effectively represents the styles of clothes based on their visual image.
Moreover, the team created an optimization model that effectively determines
the product mix for each store based on the image representation of clothes. In
the past, computers were only considered to be useful for conducting logical
calculations, and visual perception and cognition were considered to be
difficult computational tasks. The proposed approach is significant in that it
uses both AI (perception and cognition) and mathematical optimization (logical
calculation) to address a practical supply chain problem, which is why the
study was called "Breaking Moravec's Paradox."

本研究介绍了使用人工智能和优化方法结合的方式来分配时尚产品，利用深度学习模型识别产品的视觉形象，通过优化模型来决定每个商店的产品组合，以解决高品种、低容量时尚产品分配的挑战。

突破莫拉维克悖论：智能时尚零售的基于视觉的分销

Breaking Moravec's Paradox: Visual-Based Distribution in Smart Fashion  Retail

This work proposes a system for retrieving clothing and fashion products from
video content. Although films and television are the perfect showcase for
fashion brands to promote their products, spectators are not always aware of
where to buy the latest trends they see on screen. Here, a framework for
breaking the gap between fashion products shown on videos and users is
presented. By relating clothing items and video frames in an indexed database
and performing frame retrieval with temporal aggregation and fast indexing
techniques, we can find fashion products from videos in a simple and
non-intrusive way. Experiments in a large-scale dataset conducted here show
that, by using the proposed framework, memory requirements can be reduced by
42.5X with respect to linear search, whereas accuracy is maintained at around
90%.

本研究提出了一个系统，用于从视频内容中检索服装和时尚产品。通过将服装物品和视频帧存入索引数据库并使用时间聚合和快速索引技术进行帧检索，我们能够以简单和非侵入性的方式从视频中找出时尚产品。在大规模数据集上进行的实验表明，使用该框架可以将内存需求降低 42.5 倍，而精度仍然保持在约 90％左右。