Depth completion is crucial for many robotic tasks such as autonomous driving, 3-D reconstruction, and manipulation. Despite the significant progress, existing methods remain computationally intensive and often fail to meet the real-time requirements of low-power robotic platforms. Additionally, most methods are designed for opaque objects and struggle with transparent objects due to the special properties of reflection and refraction. To address these challenges, we propose a Fast Depth Completion framework for Transparent objects (FDCT), which also benefits downstream tasks like object pose estimation. To leverage local information and avoid overfitting issues when integrating it with global information, we design a new fusion branch and shortcuts to exploit low-level features and a loss function to suppress overfitting. This results in an accurate and user-friendly depth rectification framework which can recover dense depth estimation from RGB-D images alone. Extensive experiments demonstrate that FDCT can run about 70 FPS with a higher accuracy than the state-of-the-art methods. We also demonstrate that FDCT can improve pose estimation in object grasping tasks.

Current approaches for human pose estimation in videos can be categorized into per-frame and warping-based methods. Both approaches have their pros and cons. For example, per-frame methods are generally more accurate, but they are often slow. Warping-based approaches are more efficient, but the performance is usually not good. To bridge the gap, in this paper, we propose a novel fast framework for human pose estimation to meet the real-time inference with controllable accuracy degradation in compressed video domain. Our approach takes advantage of the motion representation (called ‘`motion vector’') that is readily available in a compressed video. Pose joints in a frame are obtained by directly warping the pose joints from the previous frame using the motion vectors. We also propose modules to correct possible errors introduced by the pose warping when needed. Extensive experimental results demonstrate the effectiveness of our proposed framework for accelerating the speed of top-down human pose estimation in videos.

We study an extension of lossy compression where the reconstruction is subject to a distribution constraint which can be different from the source distribution. We formulate our setting as a generalization of optimal transport with an entropy bottleneck to account for the rate constraint due to compression. We provide expressions for the tradeoff between compression rate and the achievable distortion with and without shared common randomness between the encoder and decoder. We study the examples of binary, uniform and Gaussian sources (in an asymptotic setting) in detail and demonstrate that shared randomness can strictly improve the tradeoff. For the case without common randomness and squared-Euclidean distortion, we show that the optimal solution partially decouples into the problem of optimal compression and transport and also characterize the penalty associated with fully decoupling them. We provide experimental results by training deep learning end-to-end compression systems for performing denoising on SVHN (The Street View House Numbers) and super-resolution on MNIST (Modified National Institute of Standards and Technology) datasets suggesting consistency with our theoretical results. Our code is available at https://github.com/liuh127/CrossdomainLC.

Despite their remarkable expressibility, convolution neural networks (CNNs) still fall short of delivering satisfactory results on single image dehazing, especially in terms of faithful recovery of fine texture details. In this paper, we argue that the inadequacy of conventional CNN-based dehazing methods can be attributed to the fact that the domain of hazy images is too far away from that of clear images, rendering it difficult to train a CNN for learning direct domain shift through an end-to-end manner and recovering texture details simultaneously. To address this issue, we propose to add explicit constraints inside a deep CNN model to guide the restoration process. In contrast to direct learning, the proposed mechanism shifts and narrows the candidate region for the estimation output via multiple confident neighborhoods. Therefore, it is capable of consolidating the expressibility of different architectures, resulting in a more accurate indirect domain shift (IDS) from the hazy images to that of clear images. We also propose two different training schemes, including hard IDS and soft IDS, which further reveal the effectiveness of the proposed method. Our extensive experimental results indicate that the dehazing method based on this mechanism dramatically outperforms the state-of-the-arts.

Canonical Correlation Analysis (CCA) is a classic multivariate statistical technique, which can be used to find a projection pair that maximally captures the correlation between two sets of random variables. The present paper introduces a CCA-based approach for image retrieval. It capitalizes on feature maps induced by two images under comparison through a pre-trained Convolutional Neural Network (CNN) and leverages basis vectors identified through CCA, together with an element-wise selection method based on a Chernoff-information-related criterion, to produce compact transformed image features; a binary hypothesis test regarding the joint distribution of transformed feature pair is then employed to measure the similarity between two images. The proposed approach is benchmarked against two alternative statistical methods, Linear Discriminant Analysis (LDA) and Principal Component Analysis with whitening (PCAw). Our CCA-based approach is shown to achieve highly competitive retrieval performances on standard datasets, which include, among others, Oxford5k and Paris6k.