Accurately annotating multiple 3D objects in LiDAR scenes is laborious and challenging. While a few previous studies have attempted to leverage semi-automatic methods for cost-effective bounding box annotation, such methods have limitations in efficiently handling numerous multi-class objects. To effectively accelerate 3D annotation pipelines, we propose iDet3D, an efficient interactive 3D object detector. Supporting a user-friendly 2D interface, which can ease the cognitive burden of exploring 3D space to provide click interactions, iDet3D enables users to annotate the entire objects in each scene with minimal interactions. Taking the sparse nature of 3D point clouds into account, we design a negative click simulation (NCS) to improve accuracy by reducing false-positive predictions. In addition, iDet3D incorporates two click propagation techniques to take full advantage of user interactions: (1) dense click guidance (DCG) for keeping user-provided information throughout the network and (2) spatial click propagation (SCP) for detecting other instances of the same class based on the user-specified objects. Through our extensive experiments, we present that our method can construct precise annotations in a few clicks, which shows the practicality as an efficient annotation tool for 3D object detection.

Despite the growing prevalence of black-box pre-trained models (PTMs) such as prediction API services, there remains a significant challenge in directly applying general models to real-world scenarios due to the data distribution gap. Considering a data deficiency and constrained computational resource scenario, this paper proposes a novel parameter-efficient transfer learning framework for vision recognition models in the black-box setting. Our framework incorporates two novel training techniques. First, we align the input space (i.e., image) of PTMs to the target data distribution by generating visual prompts of spatial and frequency domain. Along with the novel spatial-frequency hybrid visual prompter, we design a novel training technique based on probabilistic clusters, which can enhance class separation in the output space (i.e., prediction probabilities). In experiments, our model demonstrates superior performance in a few-shot transfer learning setting across extensive visual recognition datasets, surpassing state-of-the-art baselines. Additionally, we show that the proposed method efficiently reduces computational costs for training and inference phases.

Interactive segmentation methods have been investigated to address the potential need for additional refinement in automatic segmentation via human-in-the-loop techniques. For accurate segmentation of 3D images, we propose Slice-and-Conquer, a novel planar-to-3D pipeline formulating volumetric mask construction into two stages: 1) 2D interactive segmentation and 2) guided 3D segmentation. Specifically, the first stage enables users to focus on a single 2D slice and provides the corresponding 2D prediction results as strong shape priors. Taking the planar guidance, an accurate 3D mask can be constructed with minimal interactions. To support a flexible iterative refinement, our system recommends a next slice to annotate at the end of the second stage. Since volumetric segmentation can be completed by consecutively annotating a few recommended 2D slices, our method significantly reduces the cognitive burden of exploring volumetric space for users. Through extensive experiments on various datasets of 3D biomedical images, we demonstrate the effectiveness of the proposed pipeline.

Machine learning-based algorithms using fully convolutional networks (FCNs) have been a promising option for medical image segmentation. However, such deep networks silently fail if input samples are drawn far from the training data distribution, thus causing critical problems in automatic data processing pipelines. To overcome such out-of-distribution (OoD) problems, we propose a novel OoD score formulation and its regularization strategy by applying an auxiliary add-on classifier to an intermediate layer of an FCN, where the auxiliary module is helfpul for analyzing the encoder output features by taking their class information into account. Our regularization strategy train the module along with the FCN via the principle of outlier exposure so that our model can be trained to distinguish OoD samples from normal ones without modifying the original network architecture. Our extensive experiment results demonstrate that the proposed approach can successfully conduct effective OoD detection without loss of segmentation performance. In addition, our module can provide reasonable explanation maps along with OoD scores, which can enable users to analyze the reliability of predictions.

In this paper, we study the security of a compressed sensing (CS) based cryptosystem called a sparse one-time sensing (S-OTS) cryptosystem, which encrypts a plaintext with a sparse measurement matrix. To construct the secret matrix and renew it at each encryption, a bipolar keystream and a random permutation pattern are employed as cryptographic primitives, which can be obtained by a keystream generator of stream ciphers. With a small number of nonzero elements in the measurement matrix, the S-OTS cryptosystem achieves efficient CS encryption in terms of memory and computational cost. In security analysis, we show that the S-OTS cryptosystem can be indistinguishable as long as each plaintext has constant energy, which formalizes computational security against ciphertext only attacks (COA). In addition, we consider a chosen plaintext attack (CPA) against the S-OTS cryptosystem, which consists of two sequential stages, keystream and key recovery attacks. Against keystream recovery under CPA, we demonstrate that the S-OTS cryptosystem can be secure with overwhelmingly high probability, as an adversary needs to distinguish a prohibitively large number of candidate keystreams. Finally, we conduct an information-theoretic analysis to show that the S-OTS cryptosystem can be resistant against key recovery under CPA by guaranteeing that the probability of success is extremely low. In conclusion, the S-OTS cryptosystem can be computationally secure against COA and the two-stage CPA, while providing efficiency in CS encryption.

In this paper, we study the security of a compressed sensing (CS) based cryptosystem that encrypts a plaintext with a sparse measurement matrix. The secret matrix is constructed by a bipolar keystream and a random permutation, and renewed at each encryption. The CS-based cryptosystem performs efficient encryption with a small number of nonzero entries in the matrix, and guarantees reliable decryption for a legitimate recipient. Through a quantitative analysis, we demonstrate that the CS-based cryptosystem achieves the security against a chosen plaintext attack (CPA) with overwhelmingly high probability, by showing that an adversary needs to distinguish a prohibitively large number of candidate keystreams.

In this letter, we study the security of a cryptosystem over wireless channels that employs the asymptotically Gaussian compressed encryption. We investigate the indistinguishability and the energy sensitivity of the cryptosystem, where the total variation (TV) distance is examined as a statistical measure for the indistinguishability. To characterize the TV distance, we compute the Hellinger distance between probability distributions of ciphertexts, each of which can be modeled as a circularly symmetric complex Gaussian random vector with a constraint on plaintexts. Using the distance metrics, we show that the cryptosystem can be a promising option for secure wireless communications by guaranteeing the indistinguishability against an eavesdropper, as long as each plain text has constant energy.