Abstract: Human detection is a key component in fields such as advanced driving assistance and video surveillance. However, even detecting non-occluded standing humans remains a challenge of intensive research. Finding good features to build human models for further detection is probably one of the most important issues to face. Currently, shape, texture and motion features have deserve extensive attention in the literature. However, color-based features, which are important in other domains (e.g., image categorization), have received much less attention. In fact, the use of RGB color space has become a kind of choice by default. The focus has been put in developing first and second order features on top of RGB space (e.g., HOG and co-occurrence matrices, resp.). In this paper we evaluate the opponent colors (OPP) space as a biologically inspired alternative for human detection. In particular, by feeding OPP space in the baseline framework of Dalal et al. for human detection (based on RGB, HOG and linear SVM), we will obtain better detection performance than by using RGB space. This is a relevant result since, up to the best of our knowledge, OPP space has not been previously used for human detection. This suggests that in the future it could be worth to compute co-occurrence matrices, self-similarity features, etc., also on top of OPP space, i.e., as we have done with HOG in this paper.

Abstract: Ground-truth data is essential for the objective evaluation of object detection methods in computer vision. Many works claim their method is robust but they support it with experiments which are not quantitatively assessed with regard some ground-truth. This is one of the main obstacles to properly evaluate and compare such methods. One of the main reasons is that creating an extensive and representative ground-truth is very time consuming, specially in the case of video sequences, where thousands of frames have to be labelled. Could such a ground-truth be generated, at least in part, automatically? Though it may seem a contradictory question, we show that this is possible for the case of video sequences recorded from a moving camera. The key idea is transferring existing frame segmentations from a reference sequence into another video sequence recorded at a different time on the same track, possibly under a different ambient lighting. We have carried out experiments on several video sequence pairs and quantitatively assessed the precision of the transformed ground-truth, which prove that our approach is not only feasible but also quite accurate.

Abstract: Generative Adversarial Networks (GANs) have recently demonstrated to successfully approximate complex data distributions. A relevant extension of this model is conditional GANs (cGANs), where the introduction of external information allows to determine specific representations of the generated images. In this work, we evaluate encoders to inverse the mapping of a cGAN, i.e., mapping a real image into a latent space and a conditional representation. This allows, for example, to reconstruct and modify real images of faces conditioning on arbitrary attributes.
Additionally, we evaluate the design of cGANs. The combination of an encoder
with a cGAN, which we call Invertible cGAN (IcGAN), enables to re-generate real
images with deterministic complex modifications.

Abstract: Existing state of the art optical flow approaches, which are evaluated on standard datasets such as Middlebury, not necessarily have a similar performance when evaluated on driving scenarios. This drop on performance is due to several challenges arising on real scenarios during driving. Towards this direction, in this paper, we propose a modification to the regularization term in a variational optical flow formulation, that notably improves the results, specially in driving scenarios. The proposed modification consists on using the Laplacian derivatives of flow components in the regularization term instead of gradients of flow components. We show the improvements in results on a standard real image sequences dataset (KITTI).

Abstract: Driving requires reacting to a wide variety of complex environment conditions and agent behaviors. Explicitly modeling each possible scenario is unrealistic. In contrast, imitation learning can, in theory, leverage data from large fleets of human-driven cars. Behavior cloning in particular has been successfully used to learn simple visuomotor policies end-to-end, but scaling to the full spectrum of driving behaviors remains an unsolved problem. In this paper, we propose a new benchmark to experimentally investigate the scalability and limitations of behavior cloning. We show that behavior cloning leads to state-of-the-art results, executing complex lateral and longitudinal maneuvers, even in unseen environments, without being explicitly programmed to do so. However, we confirm some limitations of the behavior cloning approach: some well-known limitations (eg, dataset bias and overfitting), new generalization issues (eg, dynamic objects and the lack of a causal modeling), and training instabilities, all requiring further research before behavior cloning can graduate to real-world driving. The code, dataset, benchmark, and agent studied in this paper can be found at github.