This paper study the reconstruction of High Dynamic Range (HDR) video from snapshot-coded LDR video. Constructing an HDR video requires restoring the HDR values for each frame and maintaining the consistency between successive frames. HDR image acquisition from single image capture, also known as snapshot HDR imaging, can be achieved in several ways. For example, the reconfigurable snapshot HDR camera is realized by introducing an optical element into the optical stack of the camera; by placing a coded mask at a small standoff distance in front of the sensor. High-quality HDR image can be recovered from the captured coded image using deep learning methods. This study utilizes 3D-CNNs to perform a joint demosaicking, denoising, and HDR video reconstruction from coded LDR video. We enforce more temporally consistent HDR video reconstruction by introducing a temporal loss function that considers the short-term and long-term consistency. The obtained results are promising and could lead to affordable HDR video capture using conventional cameras.
Access to 3D point cloud representations has been widely facilitated by LiDAR sensors embedded in various mobile devices. This has led to an emerging need for fast and accurate point cloud processing techniques. In this paper, we revisit and dive deeper into PointNet++, one of the most influential yet under-explored networks, and develop faster and more accurate variants of the model. We first present a novel Separable Set Abstraction (SA) module that disentangles the vanilla SA module used in PointNet++ into two separate learning stages: (1) learning channel correlation and (2) learning spatial correlation. The Separable SA module is significantly faster than the vanilla version, yet it achieves comparable performance. We then introduce a new Anisotropic Reduction function into our Separable SA module and propose an Anisotropic Separable SA (ASSA) module that substantially increases the network's accuracy. We later replace the vanilla SA modules in PointNet++ with the proposed ASSA module, and denote the modified network as ASSANet. Extensive experiments on point cloud classification, semantic segmentation, and part segmentation show that ASSANet outperforms PointNet++ and other methods, achieving much higher accuracy and faster speeds. In particular, ASSANet outperforms PointNet++ by $7.4$ mIoU on S3DIS Area 5, while maintaining $1.6 \times $ faster inference speed on a single NVIDIA 2080Ti GPU. Our scaled ASSANet variant achieves $66.8$ mIoU and outperforms KPConv, while being more than $54 \times$ faster.
Understanding movies and their structural patterns is a crucial task to decode the craft of video editing. While previous works have developed tools for general analysis such as detecting characters or recognizing cinematography properties at the shot level, less effort has been devoted to understanding the most basic video edit, the Cut. This paper introduces the cut type recognition task, which requires modeling of multi-modal information. To ignite research in the new task, we construct a large-scale dataset called MovieCuts, which contains more than 170K videoclips labeled among ten cut types. We benchmark a series of audio-visual approaches, including some that deal with the problem's multi-modal and multi-label nature. Our best model achieves 45.7% mAP, which suggests that the task is challenging and that attaining highly accurate cut type recognition is an open research problem.
Video content creation keeps growing at an incredible pace; yet, creating engaging stories remains challenging and requires non-trivial video editing expertise. Many video editing components are astonishingly hard to automate primarily due to the lack of raw video materials. This paper focuses on a new task for computational video editing, namely the task of raking cut plausibility. Our key idea is to leverage content that has already been edited to learn fine-grained audiovisual patterns that trigger cuts. To do this, we first collected a data source of more than 10K videos, from which we extract more than 255K cuts. We devise a model that learns to discriminate between real and artificial cuts via contrastive learning. We set up a new task and a set of baselines to benchmark video cut generation. We observe that our proposed model outperforms the baselines by large margins. To demonstrate our model in real-world applications, we conduct human studies in a collection of unedited videos. The results show that our model does a better job at cutting than random and alternative baselines.
Deep learning models are prone to being fooled by imperceptible perturbations known as adversarial attacks. In this work, we study how equipping models with Test-time Transformation Ensembling (TTE) can work as a reliable defense against such attacks. While transforming the input data, both at train and test times, is known to enhance model performance, its effects on adversarial robustness have not been studied. Here, we present a comprehensive empirical study of the impact of TTE, in the form of widely-used image transforms, on adversarial robustness. We show that TTE consistently improves model robustness against a variety of powerful attacks without any need for re-training, and that this improvement comes at virtually no trade-off with accuracy on clean samples. Finally, we show that the benefits of TTE transfer even to the certified robustness domain, in which TTE provides sizable and consistent improvements.
Deep neural networks are vulnerable to small input perturbations known as adversarial attacks. Inspired by the fact that these adversaries are constructed by iteratively minimizing the confidence of a network for the true class label, we propose the anti-adversary layer, aimed at countering this effect. In particular, our layer generates an input perturbation in the opposite direction of the adversarial one, and feeds the classifier a perturbed version of the input. Our approach is training-free and theoretically supported. We verify the effectiveness of our approach by combining our layer with both nominally and robustly trained models, and conduct large scale experiments from black-box to adaptive attacks on CIFAR10, CIFAR100 and ImageNet. Our anti-adversary layer significantly enhances model robustness while coming at no cost on clean accuracy.
Active speaker detection requires a solid integration of multi-modal cues. While individual modalities can approximate a solution, accurate predictions can only be achieved by explicitly fusing the audio and visual features and modeling their temporal progression. Despite its inherent muti-modal nature, current methods still focus on modeling and fusing short-term audiovisual features for individual speakers, often at frame level. In this paper we present a novel approach to active speaker detection that directly addresses the multi-modal nature of the problem, and provides a straightforward strategy where independent visual features from potential speakers in the scene are assigned to a previously detected speech event. Our experiments show that, an small graph data structure built from a single frame, allows to approximate an instantaneous audio-visual assignment problem. Moreover, the temporal extension of this initial graph achieves a new state-of-the-art on the AVA-ActiveSpeaker dataset with a mAP of 88.8\%.
We introduce the idea of using learnable neighbor-to-grid soft assignment in grid-based aggregation functions for the task of 3D semantic segmentation. Previous methods in literature operate on a predefined geometric grid such as local volume partitions or irregular kernel points. These methods use geometric functions to assign local neighbors to their corresponding grid. Such geometric heuristics are potentially sub-optimal for the end task of semantic segmentation. Furthermore, they are applied uniformly throughout the depth of the network. A more general alternative would allow the network to learn its own neighbor-to-grid assignment function that best suits the end task. Since it is learnable, this mapping has the flexibility to be different per layer. This paper leverages learned neighbor-to-grid soft assignment to define an aggregation function that balances efficiency and performance. We demonstrate the efficacy of our method by reaching state-of-the-art (SOTA) performance on S3DIS with almost 10$\times$ less parameters than the current reigning method. We also demonstrate competitive performance on ScanNet and PartNet as compared with much larger SOTA models.
Temporal action localization (TAL) in videos is a challenging task, especially due to the large scale variation of actions. In the data, short actions usually occupy the major proportion, but have the lowest performance with all current methods. In this paper, we confront the challenge of short actions and propose a multi-level cross-scale solution dubbed as video self-stitching graph network (VSGN). We have two key components in VSGN: video self-stitching (VSS) and cross-scale graph pyramid network (xGPN). In VSS, we focus on a short period of a video and magnify it along the temporal dimension to obtain a larger scale. By our self-stitching approach, we are able to utilize the original clip and its magnified counterpart in one input sequence to take advantage of the complementary properties of both scales. The xGPN component further exploits the cross-scale correlations by a pyramid of cross-scale graph networks, each containing a hybrid temporal-graph module to aggregate features from across scales as well as within the same scale. Our VSGN not only enhances the feature representations, but also generates more positive anchors for short actions and more short training samples. Experiments demonstrate that VSGN obviously improves the localization performance of short actions as well as achieving the state-of-the-art overall performance on ActivityNet-v1.3, reaching an average mAP of 35.07 %.
Multi-view projection methods have shown the capability to reach state-of-the-art performance on 3D shape recognition. Most advances in multi-view representation focus on pooling techniques that learn to aggregate information from the different views, which tend to be heuristically set and fixed for all shapes. To circumvent the lack of dynamism of current multi-view methods, we propose to learn those viewpoints. In particular, we introduce a Multi-View Transformation Network (MVTN) that regresses optimal viewpoints for 3D shape recognition. By leveraging advances in differentiable rendering, our MVTN is trained end-to-end with any multi-view network and optimized for 3D shape classification. We show that MVTN can be seamlessly integrated into various multi-view approaches to exhibit clear performance gains in the tasks of 3D shape classification and shape retrieval without any extra training supervision. Furthermore, our MVTN improves multi-view networks to achieve state-of-the-art performance in rotation robustness and in object shape retrieval on ModelNet40.