TUD Logo

TUD Startseite » ... » Fakultät Informatik » Forschung » Wissenschaftliche Vorträge

Fakultät Informatik

Wissenschaftliche Vorträge

Implementation of the Platform Integration for the Backend of the Neural Network Inference Engine Generator FINN Emphasizing the Efficient Execution of Multi-Layer Offload Schedules

Präsentation der Diplomarbeit (alle Studiengänge) von Björn Gottschall (Institut für Technische Informatik, Professur Prozessordesign)

18.6.2018, 10:00 Uhr, BAR III81/82 (Meetingraum 3. Etage)

Artificial intelligence and machine learning has become an important research domain and is already found in many applications of everyday life, like speech recognition and synthesis for personal assistants, autonomous driving, object and face detection, translations and many more. Neural networks are playing a special role in this context, as they can be designed and trained to solve formerly intractable problems, especially in image classification and object detection.

This work proposes a platform integration for neural network inference, which is highly optimized and offers a uniform code base, that can easily be adapted to new network types and hardware platforms. The special focus are Multilayer Offload networks, which are designed to run with limited hardware resources and pose unprecedented demands to an interface driving neural networks. Highly quantized or even binarized neural networks are the key requirement to enable reprogrammable hardware to compete in inference processes with ASCIs or GPUs. DoReFaNet and Tinier-YOLO, two reduced precision Multilayer Offload networks, are currently covered by the new platform integration and potentially every convolutional network from FINN, a generator for neural network inference accelerators. Special layer implementations, which are not covered by hardware, and memory reorganization on bit level offered many optimization potentials compared to old specialized interface implementations. Together with batch processing on the interface side and support for multi-threading, the new unified platform integration increased performance drastically for each covered network.

Diese Veranstaltung wird unterstützt von Professur Prozessordesign.

From Linear Temporal Logic to Unambigious Büchi Automata

Präsentation der Master-Thesis von Simon Jantsch

22.6.2018, 13:45 Uhr, APB 1096

Designing Accessible User Interfaces for All by Means of Adaptive Systems

Verteidigung im Promotionsverfahren von Dipl.-Medieninf. Claudia Loitsch (Institut für Angewandte Informatik, Professur Mensch-Computer Interaktion)

25.6.2018, 10:15 Uhr, APB 1096

Proposal of a Hybrid Algorithm for Burst Transmission in Wireless Sensor Networks

Verteidigung im Promotionsverfahren von M. Sc. Zeeshan Ansar (Institut für Systemarchitektur, Lehrstuhl Rechnernetze)

25.6.2018, 13:00 Uhr, APB 3105

Formal Methods for Probabilistic Energy Models

Verteidigung im Promotionsverfahren von Dipl.-Inf. Marcus Daum (Institut für Theoretische Informatik, Algebraische und logische Grundlagen der Informatik)

27.6.2018, 14:15 Uhr, APB 1096

Suche im Ankündigungsarchiv

Abonnieren Sie die Vortragsankündigungen als News Feed: RSS

Stand: 18.6.2018, 8:00 Uhr
Autor: Webmaster