Monocular depth estimation is critical for endoscopists to perform spatial perception and 3D navigation of surgical sites. However, most of the existing methods ignore the important geometric structural consistency, which inevitably leads to performance degradation and distortion of 3D reconstruction. To address this issue, we introduce a gradient loss to penalize edge fluctuations ambiguous around stepped edge structures and a normal loss to explicitly express the sensitivity to frequently small structures, and propose a geometric consistency loss to spreads the spatial information across the sample grids to constrain the global geometric anatomy structures. In addition, we develop a synthetic RGB-Depth dataset that captures the anatomical structures under reflections and illumination variations. The proposed method is extensively validated across different datasets and clinical images and achieves mean RMSE values of 0.066 (stomach), 0.029 (small intestine), and 0.139 (colon) on the EndoSLAM dataset. The generalizability of the proposed method achieves mean RMSE values of 12.604 (T1-L1), 9.930 (T2-L2), and 13.893 (T3-L3) on the ColonDepth dataset. The experimental results show that our method exceeds previous state-of-the-art competitors and generates more consistent depth maps and reasonable anatomical structures. The quality of intraoperative 3D structure perception from endoscopic videos of the proposed method meets the accuracy requirements of video-CT registration algorithms for endoscopic navigation. The dataset and the source code will be available at https://github.com/YYM-SIA/LINGMI-MR.
Passage retrieval aims to retrieve relevant passages from large collections of the open-domain corpus. Contextual Masked Auto-Encoding has been proven effective in representation bottleneck pre-training of a monolithic dual-encoder for passage retrieval. Siamese or fully separated dual-encoders are often adopted as basic retrieval architecture in the pre-training and fine-tuning stages for encoding queries and passages into their latent embedding spaces. However, simply sharing or separating the parameters of the dual-encoder results in an imbalanced discrimination of the embedding spaces. In this work, we propose to pre-train Contextual Masked Auto-Encoder with Mixture-of-Textual-Experts (CoT-MoTE). Specifically, we incorporate textual-specific experts for individually encoding the distinct properties of queries and passages. Meanwhile, a shared self-attention layer is still kept for unified attention modeling. Results on large-scale passage retrieval benchmarks show steady improvement in retrieval performances. The quantitive analysis also shows a more balanced discrimination of the latent embedding spaces.
Dexterous in-hand manipulation for a multi-fingered anthropomorphic hand is extremely difficult because of the high-dimensional state and action spaces, rich contact patterns between the fingers and objects. Even though deep reinforcement learning has made moderate progress and demonstrated its strong potential for manipulation, it is still faced with certain challenges, such as large-scale data collection and high sample complexity. Especially, for some slight change scenes, it always needs to re-collect vast amounts of data and carry out numerous iterations of fine-tuning. Remarkably, humans can quickly transfer learned manipulation skills to different scenarios with little supervision. Inspired by human flexible transfer learning capability, we propose a novel dexterous in-hand manipulation progressive transfer learning framework (PTL) based on efficiently utilizing the collected trajectories and the source-trained dynamics model. This framework adopts progressive neural networks for dynamics model transfer learning on samples selected by a new samples selection method based on dynamics properties, rewards and scores of the trajectories. Experimental results on contact-rich anthropomorphic hand manipulation tasks show that our method can efficiently and effectively learn in-hand manipulation skills with a few online attempts and adjustment learning under the new scene. Compared to learning from scratch, our method can reduce training time costs by 95%.
Growing techniques have been emerging to improve the performance of passage retrieval. As an effective representation bottleneck pretraining technique, the contextual masked auto-encoder utilizes contextual embedding to assist in the reconstruction of passages. However, it only uses a single auto-encoding pre-task for dense representation pre-training. This study brings multi-view modeling to the contextual masked auto-encoder. Firstly, multi-view representation utilizes both dense and sparse vectors as multi-view representations, aiming to capture sentence semantics from different aspects. Moreover, multiview decoding paradigm utilizes both autoencoding and auto-regressive decoders in representation bottleneck pre-training, aiming to provide both reconstructive and generative signals for better contextual representation pretraining. We refer to this multi-view pretraining method as CoT-MAE v2. Through extensive experiments, we show that CoT-MAE v2 is effective and robust on large-scale passage retrieval benchmarks and out-of-domain zero-shot benchmarks.
This article investigates the challenge of achieving functional tool-use grasping with high-DoF anthropomorphic hands, with the aim of enabling anthropomorphic hands to perform tasks that require human-like manipulation and tool-use. However, accomplishing human-like grasping in real robots present many challenges, including obtaining diverse functional grasps for a wide variety of objects, handling generalization ability for kinematically diverse robot hands and precisely completing object shapes from a single-view perception. To tackle these challenges, we propose a six-step grasp synthesis algorithm based on fine-grained contact modeling that generates physically plausible and human-like functional grasps for category-level objects with minimal human demonstrations. With the contact-based optimization and learned dense shape correspondence, the proposed algorithm is adaptable to various objects in same category and a board range of robot hand models. To further demonstrate the robustness of the framework, over 10K functional grasps are synthesized to train our neural network, named DexFG-Net, which generates diverse sets of human-like functional grasps based on the reconstructed object model produced by a shape completion module. The proposed framework is extensively validated in simulation and on a real robot platform. Simulation experiments demonstrate that our method outperforms baseline methods by a large margin in terms of grasp functionality and success rate. Real robot experiments show that our method achieved an overall success rate of 79\% and 68\% for tool-use grasp on 3-D printed and real test objects, respectively, using a 5-Finger Schunk Hand. The experimental results indicate a step towards human-like grasping with anthropomorphic hands.
This paper presents a novel grid-based NeRF called F2-NeRF (Fast-Free-NeRF) for novel view synthesis, which enables arbitrary input camera trajectories and only costs a few minutes for training. Existing fast grid-based NeRF training frameworks, like Instant-NGP, Plenoxels, DVGO, or TensoRF, are mainly designed for bounded scenes and rely on space warping to handle unbounded scenes. Existing two widely-used space-warping methods are only designed for the forward-facing trajectory or the 360-degree object-centric trajectory but cannot process arbitrary trajectories. In this paper, we delve deep into the mechanism of space warping to handle unbounded scenes. Based on our analysis, we further propose a novel space-warping method called perspective warping, which allows us to handle arbitrary trajectories in the grid-based NeRF framework. Extensive experiments demonstrate that F2-NeRF is able to use the same perspective warping to render high-quality images on two standard datasets and a new free trajectory dataset collected by us. Project page: https://totoro97.github.io/projects/f2-nerf.
Object picking in cluttered scenes is a widely investigated field of robot manipulation, however, ambidextrous robot picking is still an important and challenging issue. We found the fusion of different prehensile actions (grasp and suction) can expand the range of objects that can be picked by robot, and the fusion of prehensile action and nonprehensile action (push) can expand the picking space of ambidextrous robot. In this paper, we propose a Push-Grasp-Suction (PGS) tri-mode grasping learning network for ambidextrous robot picking through the fusion of different prehensile actions and the fusion of prehensile action and nonprehensile aciton. The prehensile branch of PGS takes point clouds as input, and the 6-DoF picking configuration of grasp and suction in cluttered scenes are generated by multi-task point cloud learning. The nonprehensile branch with depth image input generates instance segmentation map and push configuration, cooperating with the prehensile actions to complete the picking of objects out of single-arm space. PGS generalizes well in real scene and achieves state-of-the-art picking performance.
In Vehicle-to-Everything (V2X) communication, the high mobility of vehicles generates the Doppler shift which leads to channel uncertainties. Moreover, the reasons for channel uncertainties also include the finite channel feedback, channels state information (CSI) loss and latency. With this concern, we formulate a joint spectrum and power allocation problem for V2X communication with imperfect CSI. Specifically, the sum capacity of cellular user equipments (CUEs) is maximized subject to the minimum Signal-to-Interference-and-Noise Ratio (SINR) requirements of CUEs and the outage probability constraints of vehicular user equipments (VUEs). Then, two different robust resource allocation approaches are designed to solve the problem. One is Bernstein Approximation-based Robust Resource Allocation approach. More specifically, Bernstein approximations are employed to convert the chance constraint into a calculable constraint, and Bisection search method is proposed to obtain the optimal allocation solution with low complexity. Then, for further reducing the computational complexity, Self-learning Robust Resource Allocation approach, which includes a learning method and an analytical mapping method, is proposed as the second approach. The learning method is devised to learn the uncertainty set which transforms the chance constraint into calculable constraints, and the analytical mapping method is proposed to obtain closed-form solutions of the resource allocation problem. Finally, the simulation results prove that the proposed approaches can improve the capacity of all CUEs effectively whilst ensuring the reliability of the channel.
Object picking in cluttered scenes is a widely investigated field of robot manipulation, however, ambidextrous robot picking is still an important and challenging issue. We found the fusion of different prehensile actions (grasp and suction) can expand the range of objects that can be picked by robot, and the fusion of prehensile action and nonprehensile action (push) can expand the picking space of ambidextrous robot. In this paper, we propose a "Push-Grasp-Suction" (PGS) integrated network for ambidextrous robot picking through the fusion of different prehensile actions and the fusion of prehensile action and nonprehensile aciton. The prehensile branch of PGS takes point clouds as input, and the 6-DoF picking configuration of grasp and suction in cluttered scenes are generated by multi-task point cloud learning. The nonprehensile branch with depth image input generates instance segmentation map and push configuration, cooperating with the prehensile actions to complete the picking of objects out of single-arm space. PGS generalizes well in real scene and achieves state-of-the-art picking performance.
Self-supervised node representation learning aims to learn node representations from unlabelled graphs that rival the supervised counterparts. The key towards learning informative node representations lies in how to effectively gain contextual information from the graph structure. In this work, we present simple-yet-effective self-supervised node representation learning via aligning the hidden representations of nodes and their neighbourhood. Our first idea achieves such node-to-neighbourhood alignment by directly maximizing the mutual information between their representations, which, we prove theoretically, plays the role of graph smoothing. Our framework is optimized via a surrogate contrastive loss and a Topology-Aware Positive Sampling (TAPS) strategy is proposed to sample positives by considering the structural dependencies between nodes, which enables offline positive selection. Considering the excessive memory overheads of contrastive learning, we further propose a negative-free solution, where the main contribution is a Graph Signal Decorrelation (GSD) constraint to avoid representation collapse and over-smoothing. The GSD constraint unifies some of the existing constraints and can be used to derive new implementations to combat representation collapse. By applying our methods on top of simple MLP-based node representation encoders, we learn node representations that achieve promising node classification performance on a set of graph-structured datasets from small- to large-scale.