Enviar aquest missatge de text: Learning Cross-Modal Embeddings with Adversarial Networks for Cooking Recipes and Food Images