CoCar NextGen: a Multi-Purpose Platform for Connected Autonomous Driving Research

Real world testing is of vital importance to the success of automated driving. While many players in the business design purpose build testing vehicles, we designed and build a modular platform that offers high flexibility for any kind of scenario. CoCar NextGen is equipped with next generation hardware that addresses all future use cases. Its extensive, redundant sensor setup allows to develop cross-domain data driven approaches that manage the transfer to other sensor setups. Together with the possibility of being deployed on public roads, this creates a unique research platform that supports the road to automated driving on SAE Level 5.

Balancing Expert Utilization in Mixture-of-Experts Layers Embedded in CNNs

This work addresses the problem of unbalanced expert utilization in sparsely-gated Mixture of Expert (MoE) layers, embedded directly into convolutional neural networks. To enable a stable training process, we present both soft and hard constraint-based approaches. With hard constraints, the weights of certain experts are allowed to become zero, while soft constraints balance the contribution of experts with an additional auxiliary loss. As a result, soft constraints handle expert utilization better and support the expert specialization process, hard constraints mostly maintain generalized experts and increase the model performance for many applications. Our findings demonstrate that even with a single dataset and end-to-end training, experts can implicitly focus on individual sub-domains of the input space. Experts in the proposed models with MoE embeddings implicitly focus on distinct domains, even without suitable predefined datasets. As an example, experts trained for CIFAR-100 image classification specialize in recognizing different domains such as sea animals or flowers without previous data clustering. Experiments with RetinaNet and the COCO dataset further indicate that object detection experts can also specialize in detecting objects of distinct sizes.