Topic outline

  • These are the seminar topics for this year. The seminar presentations schedule is in the general course description, here.

    Please select the topic by replying with your name, email, and topic number to the post entitled "TOPIC SELECTION" in the announcements, here.


    The seminar topics are:

    Topic 1: In-memory computing using resistance-based memory elements, such as ReRAM or PCM.

    Topic 2: In-memory computing using charge-based memory elements, such as SRAM or DRAM. 

    Topic 3: Comparison of three classic convolutional neural networks, such as LeNet, ImageNet (AlexNet), and VGG 16.

    Topic 4: Comparison of three modern convolutional neural networks, such as Inception (GoogLeNet), ResNet, and DenseNet. 

    Topic 5: In-memory computing applications, such as scientific computing, signal processing, and machine learning.

    Topic 6: comparison of most used activation functions in deep neural networks and their circuit realizations in analog and digital neural networks.

    Topic 7: Study on analog memories to exclude data conversion in neural networks.

    Topic 8: Study on summing amplifiers both in digital and analog neural networks.