In this work, we present the depth-adaptive deep neural network using a depth map for semantic segmentation. Typical deep neural networks receive inputs at the predetermined locations regardless of the distance from the camera. This fixed receptive field presents a challenge to generalize the features of objects at various distances in neural networks. Specifically, the predetermined receptive fields are too small at a short distance, and vice versa. To overcome this challenge, we develop a neural network which is able to adapt the receptive field not only for each layer but also for each neuron at spatial locations. To adjust the receptive field, we propose the depth-adaptive multiscale (DaM) convolution layer consisting of the adaptive perception neuron and the in-layer multiscale neuron. The adaptive perception neuron is to adjust the receptive field at each spatial location using the corresponding depth information. The in-layer multiscale neuron is to apply the different size of the receptive field at each feature space to learn features at multiple scales. The proposed DaM convolution is applied to two fully convolutional neural networks. We demonstrate the effectiveness of the proposed neural networks on the publicly available RGB-D dataset for semantic segmentation and the novel hand segmentation dataset for hand-object interaction. The experimental results show that the proposed method outperforms the state-of-the-art methods without any additional layers or pre/post-processing.
14 March 2019 in IEEE Transactions on Image Processing
We present a random forest framework that learns the weights, shapes, and sparsities of feature representations for real-time semantic segmentation. Typical filters (kernels) have predetermined shapes and sparsities and learn only weights. A few feature extraction methods fix weights and learn only shapes and sparsities. These predetermined constraints restrict learning and extracting optimal features. To overcome this limitation, we propose an unconstrained representation that is able to extract optimal features by learning weights, shapes, and sparsities. We, then, present the random forest framework that learns the flexible filters using an iterative optimization algorithm and segments input images using the learned representations. We demonstrate the effectiveness of the proposed method using a hand segmentation dataset for hand-object interaction and using two semantic segmentation datasets. The results show that the proposed method achieves real-time semantic segmentation using limited computational and memory resources.
24 October 2023 in Engineering Applications of Artificial Intelligence
Person re-identification is a problem of identifying individuals across non-overlapping cameras. Although remarkable progress has been made in the re-identification problem, it is still a challenging problem due to appearance variations of the same person as well as other people of similar appearance. Some prior works solved the issues by separating features of positive samples from features of negative ones. However, the performances of existing models considerably depend on the characteristics and statistics of the samples used for training. Thus, we propose a novel framework named sampling independent robust feature representation network (SirNet) that learns disentangled feature embedding from randomly chosen samples. A carefully designed sampling independent maximum discrepancy loss is introduced to model samples of the same person as a cluster. As a result, the proposed framework can generate additional hard negatives/positives using the learned features, which results in better discriminability from other identities. Extensive experimental results on large-scale benchmark datasets verify that the proposed model is more effective than prior state-of-the-art models.
Driving is a task that puts heavy demands on visual information, thereby the human visual system plays a critical role in making proper decisions for safe driving. Understanding a driver’s visual attention and relevant behavior information is a challenging but essential task in advanced driver-assistance systems (ADAS) and efficient autonomous vehicles (AV). Specifically, robust prediction of a driver’s attention from images could be a crucial key to assist intelligent vehicle systems where a self-driving car is required to move safely interacting with the surrounding environment. Thus, in this paper, we investigate a human driver’s visual behavior in terms of computer vision to estimate the driver’s attention locations in images. First, we show that feature representations at high resolution improves visual attention prediction accuracy and localization performance when being fused with features at low-resolution. To demonstrate this, we employ a deep convolutional neural network framework that learns and extracts feature representations at multiple resolutions. In particular, the network maintains the feature representation with the highest resolution at the original image resolution. Second, attention prediction tends to be biased toward centers of images when neural networks are trained using typical visual attention datasets. To avoid overfitting to the center-biased solution, the network is trained using diverse regions of images. Finally, the experimental results verify that our proposed framework improves the prediction accuracy of a driver’s attention locations.
7 February 2022 in IEEE Robotics and Automation Letters
This work addresses the task of electric vehicle (EV) charging inlet detection for autonomous EV charging robots. Recently, automated EV charging systems have received huge attention to improve users’ experience and to efficiently utilize charging infrastructures and parking lots. However, most related works have focused on system design, robot control, planning, and manipulation. Towards robust EV charging inlet detection, we propose a new dataset (EVCI dataset) and a novel data augmentation method that is based on image-to-image translation where typical image-to-image translation methods synthesize a new image in a different domain given an image. To the best of our knowledge, the EVCI dataset is the first EV charging inlet dataset. For the data augmentation method, we focus on being able to control synthesized images’ captured environments (e.g., time, lighting) in an intuitive way. To achieve this, we first propose the environment guide vector that humans can intuitively interpret. We then propose a novel image-to-image translation network that translates a given image towards the environment described by the vector. Accordingly, it aims to synthesize a new image that has the same content as the given image while looking like captured in the provided environment by the environment guide vector. Lastly, we train a detection method using the augmented dataset. Through experiments on the EVCI dataset, we demonstrate that the proposed method outperforms the state-of-the-art methods. We also show that the proposed method is able to control synthesized images using an image and environment guide vectors.