Min Li

Alex Jinpeng Wang, Dongxing Mao, Jiawei Zhang et al. · microsoft-research

Text-conditioned image generation has gained significant attention in recent years and are processing increasingly longer and comprehensive text prompt. In everyday life, dense and intricate text appears in contexts like advertisements, infographics, and signage, where the integration of both text and visuals is essential for conveying complex information. However, despite these advances, the generation of images containing long-form text remains a persistent challenge, largely due to the limitations of existing datasets, which often focus on shorter and simpler text. To address this gap, we introduce TextAtlas5M, a novel dataset specifically designed to evaluate long-text rendering in text-conditioned image generation. Our dataset consists of 5 million long-text generated and collected images across diverse data types, enabling comprehensive evaluation of large-scale generative models on long-text image generation. We further curate 3000 human-improved test set TextAtlasEval across 3 data domains, establishing one of the most extensive benchmarks for text-conditioned generation. Evaluations suggest that the TextAtlasEval benchmarks present significant challenges even for the most advanced proprietary models (e.g. GPT4o with DallE-3), while their open-source counterparts show an even larger performance gap. These evidences position TextAtlas5M as a valuable dataset for training and evaluating future-generation text-conditioned image generation models.

Ge Yunfei, Diao Changyu, Li Min et al.

The authenticity of calligraphy is significant but difficult task in the realm of art, where the key problem is the few-shot classification of calligraphy. We propose a novel method, ShufaNet ("Shufa" is the pinyin of Chinese calligraphy), to classify Chinese calligraphers' styles based on metric learning in the case of few-shot, whose classification accuracy exceeds the level of students majoring in calligraphy. We present a new network architecture, including the unique expression of the style of handwriting fonts called ShufaLoss and the calligraphy category information as prior knowledge. Meanwhile, we modify the spatial attention module and create ShufaAttention for handwriting fonts based on the traditional Chinese nine Palace thought. For the training of the model, we build a calligraphers' data set. Our method achieved 65% accuracy rate in our data set for few-shot learning, surpassing resNet and other mainstream CNNs. Meanwhile, we conducted battle for calligraphy major students, and finally surpassed them. This is the first attempt of deep learning in the field of calligrapher classification, and we expect to provide ideas for subsequent research.

Xiongshi Deng, Min Li, Shaobo Deng et al.

Microarray gene expression data are often accompanied by a large number of genes and a small number of samples. However, only a few of these genes are relevant to cancer, resulting in signigicant gene selection challenges. Hence, we propose a two-stage gene selection approach by combining extreme gradient boosting (XGBoost) and a multi-objective optimization genetic algorithm (XGBoost-MOGA) for cancer classification in microarray datasets. In the first stage, the genes are ranked use an ensemble-based feature selection using XGBoost. This stage can effectively remove irrelevant genes and yield a group comprising the most relevant genes related to the class. In the second stage, XGBoost-MOGA searches for an optimal gene subset based on the most relevant genes's group using a multi-objective optimization genetic algorithm. We performed comprehensive experiments to compare XGBoost-MOGA with other state-of-the-art feature selection methods using two well-known learning classifiers on 13 publicly available microarray expression datasets. The experimental results show that XGBoost-MOGA yields significantly better results than previous state-of-the-art algorithms in terms of various evaluation criteria, such as accuracy, F-score, precision, and recall.