Jacob Huh Rose Hills
Unsupervsied Deep Multimodal Learning for Object Detection
Recent advances in the artificial neural networks have shown immense success in the computer vision field. While supervised learning tasks in object detection has shown to be a feasible task, the necessity of the supervision in the task is not observed in humans. Despite the fact that object Information is a unison of all senses, currently, object detection tasks are restricted to visual data. I will be researching on using multi modal data to tackle object detection in a supervised and unsupervised setting.