Affordance detection is a challenging problem with a wide variety of robotic applications. Traditional affordance detection methods are limited to a predefined set of affordance labels, hence potentially restricting the adaptability of intelligent robots in complex and dynamic environments. In this paper, we present the Open-Vocabulary Affordance Detection (OpenAD) method, which is capable of detecting an unbounded number of affordances in 3D point clouds. By simultaneously learning the affordance text and the point feature, OpenAD successfully exploits the semantic relationships between affordances. Therefore, our proposed method enables zero-shot detection and can detect previously unseen affordances without a single annotation example. Intensive experimental results show that OpenAD works effectively on a wide range of affordance detection setups and outperforms other baselines by a large margin. Additionally, we demonstrate the practicality of the proposed OpenAD in real-world robotic applications with a fast inference speed (~100 ms).
In this paper, we present a high-performance and light-weight deep learning model for Remote Sensing Image Classification (RSIC), the task of identifying the aerial scene of a remote sensing image. To this end, we first valuate various benchmark convolutional neural network (CNN) architectures: MobileNet V1/V2, ResNet 50/151V2, InceptionV3/InceptionResNetV2, EfficientNet B0/B7, DenseNet 121/201, ConNeXt Tiny/Large. Then, the best performing models are selected to train a compact model in a teacher-student arrangement. The knowledge distillation from the teacher aims to achieve high performance with significantly reduced complexity. By conducting extensive experiments on the NWPU-RESISC45 benchmark, our proposed teacher-student models outperforms the state-of-the-art systems, and has potential to be applied on a wide rage of edge devices.
This paper presents ViDeBERTa, a new pre-trained monolingual language model for Vietnamese, with three versions - ViDeBERTa_xsmall, ViDeBERTa_base, and ViDeBERTa_large, which are pre-trained on a large-scale corpus of high-quality and diverse Vietnamese texts using DeBERTa architecture. Although many successful pre-trained language models based on Transformer have been widely proposed for the English language, there are still few pre-trained models for Vietnamese, a low-resource language, that perform good results on downstream tasks, especially Question answering. We fine-tune and evaluate our model on three important natural language downstream tasks, Part-of-speech tagging, Named-entity recognition, and Question answering. The empirical results demonstrate that ViDeBERTa with far fewer parameters surpasses the previous state-of-the-art models on multiple Vietnamese-specific natural language understanding tasks. Notably, ViDeBERTa_base with 86M parameters, which is only about 23% of PhoBERT_large with 370M parameters, still performs the same or better results than the previous state-of-the-art model. Our ViDeBERTa models are available at: https://github.com/HySonLab/ViDeBERTa.
This paper presents a two-step algorithm for online trajectory planning in indoor environments with unknown obstacles. In the first step, sampling-based path planning techniques such as the optimal Rapidly exploring Random Tree (RRT*) algorithm and the Line-of-Sight (LOS) algorithm are employed to generate a collision-free path consisting of multiple waypoints. Then, in the second step, constrained quadratic programming is utilized to compute a smooth trajectory that passes through all computed waypoints. The main contribution of this work is the development of a flexible trajectory planning framework that can detect changes in the environment, such as new obstacles, and compute alternative trajectories in real time. The proposed algorithm actively considers all changes in the environment and performs the replanning process only on waypoints that are occupied by new obstacles. This helps to reduce the computation time and realize the proposed approach in real time. The feasibility of the proposed algorithm is evaluated using the Intel Aero Ready-to-Fly (RTF) quadcopter in simulation and in a real-world experiment.
Solving the analytical inverse kinematics (IK) of redundant manipulators in real time is a difficult problem in robotics since its solution for a given target pose is not unique. Moreover, choosing the optimal IK solution with respect to application-specific demands helps to improve the robustness and to increase the success rate when driving the manipulator from its current configuration towards a desired pose. This is necessary, especially in high-dynamic tasks like catching objects in mid-flights. To compute a suitable target configuration in the joint space for a given target pose in the trajectory planning context, various factors such as travel time or manipulability must be considered. However, these factors increase the complexity of the overall problem which impedes real-time implementation. In this paper, a real-time framework to compute the analytical inverse kinematics of a redundant robot is presented. To this end, the analytical IK of the redundant manipulator is parameterized by so-called redundancy parameters, which are combined with a target pose to yield a unique IK solution. Most existing works in the literature either try to approximate the direct mapping from the desired pose of the manipulator to the solution of the IK or cluster the entire workspace to find IK solutions. In contrast, the proposed framework directly learns these redundancy parameters by using a neural network (NN) that provides the optimal IK solution with respect to the manipulability and the closeness to the current robot configuration. Monte Carlo simulations show the effectiveness of the proposed approach which is accurate and real-time capable ($\approx$ \SI{32}{\micro\second}) on the KUKA LBR iiwa 14 R820.
Despite significant progress over the past few years, ambiguity is still a key challenge in Facial Expression Recognition (FER). It can lead to noisy and inconsistent annotation, which hinders the performance of deep learning models in real-world scenarios. In this paper, we propose a new uncertainty-aware label distribution learning method to improve the robustness of deep models against uncertainty and ambiguity. We leverage neighborhood information in the valence-arousal space to adaptively construct emotion distributions for training samples. We also consider the uncertainty of provided labels when incorporating them into the label distributions. Our method can be easily integrated into a deep network to obtain more training supervision and improve recognition accuracy. Intensive experiments on several datasets under various noisy and ambiguous settings show that our method achieves competitive results and outperforms recent state-of-the-art approaches. Our code and models are available at https://github.com/minhnhatvt/label-distribution-learning-fer-tf.
The long-tailed distribution is a common phenomenon in the real world. Extracted large scale image datasets inevitably demonstrate the long-tailed property and models trained with imbalanced data can obtain high performance for the over-represented categories, but struggle for the under-represented categories, leading to biased predictions and performance degradation. To address this challenge, we propose a novel de-biasing method named Inverse Image Frequency (IIF). IIF is a multiplicative margin adjustment transformation of the logits in the classification layer of a convolutional neural network. Our method achieves stronger performance than similar works and it is especially useful for downstream tasks such as long-tailed instance segmentation as it produces fewer false positive detections. Our extensive experiments show that IIF surpasses the state of the art on many long-tailed benchmarks such as ImageNet-LT, CIFAR-LT, Places-LT and LVIS, reaching 55.8% top-1 accuracy with ResNet50 on ImageNet-LT and 26.2% segmentation AP with MaskRCNN on LVIS. Code available at https://github.com/kostas1515/iif
Depth estimation is a crucial step for image-guided intervention in robotic surgery and laparoscopic imaging system. Since per-pixel depth ground truth is difficult to acquire for laparoscopic image data, it is rarely possible to apply supervised depth estimation to surgical applications. As an alternative, self-supervised methods have been introduced to train depth estimators using only synchronized stereo image pairs. However, most recent work focused on the left-right consistency in 2D and ignored valuable inherent 3D information on the object in real world coordinates, meaning that the left-right 3D geometric structural consistency is not fully utilized. To overcome this limitation, we present M3Depth, a self-supervised depth estimator to leverage 3D geometric structural information hidden in stereo pairs while keeping monocular inference. The method also removes the influence of border regions unseen in at least one of the stereo images via masking, to enhance the correspondences between left and right images in overlapping areas. Intensive experiments show that our method outperforms previous self-supervised approaches on both a public dataset and a newly acquired dataset by a large margin, indicating a good generalization across different samples and laparoscopes.
Explaining artificial intelligence (AI) predictions is increasingly important and even imperative in many high-stakes applications where humans are the ultimate decision-makers. In this work, we propose two novel architectures of self-interpretable image classifiers that first explain, and then predict (as opposed to post-hoc explanations) by harnessing the visual correspondences between a query image and exemplars. Our models consistently improve (by 1 to 4 points) on out-of-distribution (OOD) datasets while performing marginally worse (by 1 to 2 points) on in-distribution tests than ResNet-50 and a $k$-nearest neighbor classifier (kNN). Via a large-scale, human study on ImageNet and CUB, our correspondence-based explanations are found to be more useful to users than kNN explanations. Our explanations help users more accurately reject AI's wrong decisions than all other tested methods. Interestingly, for the first time, we show that it is possible to achieve complementary human-AI team accuracy (i.e., that is higher than either AI-alone or human-alone), in ImageNet and CUB image classification tasks.
Autonomous robots in endovascular operations have the potential to navigate circulatory systems safely and reliably while decreasing the susceptibility to human errors. However, there are numerous challenges involved with the process of training such robots such as long training duration due to sample inefficiency of machine learning algorithms and safety issues arising from the interaction between the catheter and the endovascular phantom. Physics simulators have been used in the context of endovascular procedures, but they are typically employed for staff training and generally do not conform to the autonomous cannulation goal. Furthermore, most current simulators are closed-source which hinders the collaborative development of safe and reliable autonomous systems. In this work, we introduce CathSim, an open-source simulation environment that accelerates the development of machine learning algorithms for autonomous endovascular navigation. We first simulate the high-fidelity catheter and aorta with the state-of-the-art endovascular robot. We then provide the capability of real-time force sensing between the catheter and the aorta in the simulation environment. We validate our simulator by conducting two different catheterisation tasks within two primary arteries using two popular reinforcement learning algorithms, Proximal Policy Optimization (PPO) and Soft Actor-Critic (SAC). The experimental results show that using our open-source simulator, we can successfully train the reinforcement learning agents to perform different autonomous cannulation tasks.