UMUTeam at SemEval-2022 Task 5: Combining image and textual embeddings for multi-modal automatic misogyny identification