Most of the existing multimodal event classi cation methods fuse the traditional

hand-crafted features with some manually de ned weights, which may be not

suitable to the event classi cation task with large amounts of photos. Besides,

the feature extraction and event classi cation model is always performed separately,

which cannot capture the most useful features to describe the semantic

concepts of complex events. To deal with these issues, we propose a novel discriminative

multimodal embedding (DME) model for event classi cation in user

generated photos by jointly learning the representation together with the classifi er in a uni ed framework. In the proposed DME model, we can effectively

resolve the multimodal, intra-class variation and inter-class confusion challenges

by using the contrastive constraints on the multimodal event data. Extensive

experimental results on two collected datasets demonstrate the effectiveness of

the proposed DME model for event classi cation.