Most recent works on optical flow use convex upsampling as the last step to obtain high-resolution flow. In this work, we show and discuss several issues and limitations of this currently widely adopted convex upsampling approach. We propose a series of changes, inspired by the observation that convex upsampling as currently implemented performs badly in high-detail areas. We identify three possible causes; wrong training data, the non-existence of a convex combination, and the inability of the convex upsampler to find the correct convex combination. We propose several ideas in an attempt to resolve current issues. First, we propose to decouple the weights for the final convex upsampler, making it easier to find the correct convex combination. For the same reason, we also provide extra contextual features to the convex upsampler. Then, we increase the convex mask size by using an attention-based alternative convex upsampler; Transformers for Convex Upsampling. This upsampler is based on the observation that convex upsampling can be reformulated as attention, and we propose to use local attention masks as a drop-in replacement for convex masks in order to increase the mask size. We provide empirical evidence that a larger mask size increases the likelihood of the existence of the convex combination. Lastly, we propose an alternative training scheme to remove bilinear interpolation artifacts from the model output. We investigate whether an increase in accuracy can be achieved while leaving the low-resolution flow prediction architecture unchanged. Due to that, our proposed ideas could theoretically be applied to almost every current state-of-the-art optical flow architecture. On the FlyingChairs + FlyingThings3D training setting we reduce the Sintel Clean training end-point-error of GMA from 1.31 to 1.18, which is a 10% decrease caused solely by changes regarding the convex upsampler.

Almende B.V., a technologically innovative and research-oriented company, has been working on a new algorithm that optimizes routes for parcel delivery trucks. The algorithm contains novel features, like including the possible use of autonomous vehicles, that are at this moment in time not taken into account in existing route optimization algorithms and thus visualization applications. To this end and to get a more tangible overview of the algorithm’s behavior and performance, they requested to have a customized visualization tool developed. This report describes the process and results of developing such a tool. The tool is presented as a single-page application and has been partly depicted on the cover of this document. The goal of the project is to have a more clear overview of the routing algorithm’s capabilities, by showing its unique features on a map and displaying statistics on the side. In addition, comparing the algorithm to existing ones should provide added insights into the (expected) benefits of the new algorithm. The main purpose of the tool developed in this project is to show insight into the workings of the algorithm and to help with enhancing and developing the algorithm. An added side-bonus is that the tool can also be used to show the performance to various groups of interested parties.

Color is a crucial visual cue readily exploited by Convolutional Neural Networks (CNNs) for object recognition. However, CNNs struggle if there is data imbalance between color variations introduced by accidental recording conditions. Color invariance addresses this issue but does so at the cost of removing all color information, which sacrifices discriminative power. In this paper, we propose Color Equivariant Convolutions (CEConvs), a novel deep learning building block that enables shape feature sharing across the color spectrum while retaining important color information. We extend the notion of equivariance from geometric to photometric transformations by incorporating parameter sharing over hue-shifts in a neural network. We demonstrate the benefits of CEConvs in terms of downstream performance to various tasks and improved robustness to color changes, including train-test distribution shifts. Our approach can be seamlessly integrated into existing architectures, such as ResNets, and offers a promising solution for addressing color-based domain shifts in CNNs.@en