Robot models, particularly those trained with large amounts of data, have recently shown a plethora of real-world manipulation and navigation capabilities. Several independent efforts have shown that given sufficient training data in an environment, robot policies can generalize to demonstrated variations in that environment. However, needing to finetune robot models to every new environment stands in stark contrast to models in language or vision that can be deployed zero-shot for open-world problems. In this work, we present Robot Utility Models (RUMs), a framework for training and deploying zero-shot robot policies that can directly generalize to new environments without any finetuning. To create RUMs efficiently, we develop new tools to quickly collect data for mobile manipulation tasks, integrate such data into a policy with multi-modal imitation learning, and deploy policies ondevice on the Hello Robot Stretch, a cheap commodity robot, with an external mLLM verifier for retrying. We train five such utility models for opening cabinet doors, opening drawers, picking up napkins, picking up paper bags, and reorienting fallen objects. Our system, on average, achieves 90% success rate in unseen, novel environments interacting with unseen objects. Primary among our lessons are the importance of training data over training algorithm and policy class, guidance about data scaling, necessity for diverse yet high-quality demonstrations, and a recipe for robot introspection and retrying to improve performance on individual environments.

Generative modeling of complex behaviors from labeled datasets has been a longstanding problem in decision-making. Unlike language or image generation, decision-making requires modeling actions – continuous-valued vectors that are multimodal in their distribution, potentially drawn from uncurated sources, where generation errors can compound in sequential prediction. A recent class of models called Behavior Transformers (BeT) addresses this by discretizing actions using k-means clustering to capture different modes. However, k-means struggles to scale for high-dimensional action spaces or long sequences, and lacks gradient information, and thus BeT suffers in modeling long-range actions. In this work, we present Vector-Quantized Behavior Transformer (VQ-BeT), a versatile model for behavior generation that handles multimodal action prediction, conditional generation, and partial observations. VQ-BeT augments BeT by tokenizing continuous actions with a hierarchical vector quantization module. Across seven environments including simulated manipulation, autonomous driving, and robotics, VQ-BeT improves on state-of-the-art models such as BeT and Diffusion Policies. Importantly, we demonstrate VQ-BeT’s improved ability to capture behavior modes while accelerating inference speed 5× over Diffusion Policies. Videos can be found https://sjlee.cc/vq-bet/

Recognizing traversable terrain from 3D point cloud data is critical, as it directly impacts the performance of autonomous navigation in off-road environments. However, existing segmentation algorithms often struggle with challenges related to changes in data distribution, environmental specificity, and sensor variations. Moreover, when encountering sunken areas, their performance is frequently compromised, and they may even fail to recognize them. To address these challenges, we introduce B-TMS, a novel approach that performs map-wise terrain modeling and segmentation by utilizing Bayesian generalized kernel (BGK) within the graph structure known as the tri-grid field (TGF). Our experiments encompass various data distributions, ranging from single scans to partial maps, utilizing both public datasets representing urban scenes and off-road environments, and our own dataset acquired from extremely bumpy terrains. Our results demonstrate notable contributions, particularly in terms of robustness to data distribution variations, adaptability to diverse environmental conditions, and resilience against the challenges associated with parameter changes.

Reinforcement learning (RL) often faces the challenges of uninformed search problems where the agent should explore without access to the domain knowledge such as characteristics of the environment or external rewards. To tackle these challenges, this work proposes a new approach for curriculum RL called $\textbf{D}$iversify for $\textbf{D}$isagreement \& $\textbf{C}$onquer ($\textbf{D2C}$). Unlike previous curriculum learning methods, D2C requires only a few examples of desired outcomes and works in any environment, regardless of its geometry or the distribution of the desired outcome examples. The proposed method performs diversification of the goal-conditional classifiers to identify similarities between visited and desired outcome states and ensures that the classifiers disagree on states from out-of-distribution, which enables quantifying the unexplored region and designing an arbitrary goal-conditioned intrinsic reward signal in a simple and intuitive way. The proposed method then employs bipartite matching to define a curriculum learning objective that produces a sequence of well-adjusted intermediate goals, which enable the agent to automatically explore and conquer the unexplored region. We present experimental results demonstrating that D2C outperforms prior curriculum RL methods in both quantitative and qualitative aspects, even with the arbitrarily distributed desired outcome examples.

In the field of 3D perception using 3D LiDAR sensors, ground segmentation is an essential task for various purposes, such as traversable area detection and object recognition. Under these circumstances, several ground segmentation methods have been proposed. However, some limitations are still encountered. First, some ground segmentation methods require fine-tuning of parameters depending on the surroundings, which is excessively laborious and time-consuming. Moreover, even if the parameters are well adjusted, a partial under-segmentation problem can still emerge, which implies ground segmentation failures in some regions. Finally, ground segmentation methods typically fail to estimate an appropriate ground plane when the ground is above another structure, such as a retaining wall. To address these problems, we propose a robust ground segmentation method called Patchwork++, an extension of Patchwork. Patchwork++ exploits adaptive ground likelihood estimation (A-GLE) to calculate appropriate parameters adaptively based on the previous ground segmentation results. Moreover, temporal ground revert (TGR) alleviates a partial under-segmentation problem by using the temporary ground property. Also, region-wise vertical plane fitting (R-VPF) is introduced to segment the ground plane properly even if the ground is elevated with different layers. Finally, we present reflected noise removal (RNR) to eliminate virtual noise points efficiently based on the 3D LiDAR reflection model. We demonstrate the qualitative and quantitative evaluations using a SemanticKITTI dataset. Our code is available at https://github.com/url-kaist/patchwork-plusplus

Depending on weather conditions, optical satellites might not acquire image information of a region of interest (ROI). This is a major drawback in emergency disaster which require realtime monitoring and damage analysis. In particular, one of the most serious disasters, floods, always accompany clouds. As a result, there are difficulties in flood detection, i.e., water detection, using optical satellite images. While, Synthetic Aperture Radar (SAR) satellite has the advantage of acquiring images regardless of weather conditions such as cloud and rain. Therefore, we can effectively perform flood monitoring and damage analysis for the ROI through water detection using SAR satellite images. In this paper, we propose a deep learning-based water segmentation using KOrean Multi-Purpose SATellite (KOMPSAT-5) images. To efficiently develop the deep learning-based model, we create a SAR water dataset for over 3,000 sheets based on KOMPSAT-5. And We perform water segmentation using representative deep learning-based segmentation models such as Fully Convolutional Networks (FCN), U-Net, DeepUNet, and High Resolution Network (HRNet). Experimental results show that HRNet performs the highest accuracy, i.e, this model achieves more than 80% IoU (Intersection over Union).

Motion planners for mobile robots in unknown environments face the challenge of simultaneously maintaining both robustness against unmodeled uncertainties and persistent feasibility of the trajectory-finding problem. That is, while dealing with uncertainties, a motion planner must update its trajectory, adapting to the newly revealed environment in real-time; failing to do so may involve unsafe circumstances. Many existing planning algorithms guarantee these by maintaining the clearance needed to perform an emergency brake, which is itself a robust and persistently feasible maneuver. However, such maneuvers are not applicable for systems in which braking is impossible or risky, such as fixed-wing aircraft. To that end, we propose a real-time robust planner that recursively guarantees persistent feasibility without any need of braking. The planner ensures robustness against bounded uncertainties and persistent feasibility by constructing a loop of sequentially composed funnels, starting from the receding horizon local trajectory’s forward reachable set. We implement the proposed algorithm for a robotic car tracking a speed-fixed reference trajectory. The experiment results show that the proposed algorithm can be run at faster than 16 Hz, while successfully keeping the system away from entering any dead end, to maintain safety and feasibility.

Not only on earth, but also in space, robot systems are increasingly becoming essential elements in our lives such as a mobile exploration robot on Mars. Satellites are also an indispensable field in space robot systems, for example, from low-orbit satellites for self-driving vehicles to small satellites launched for various purposes. A lot of research is being conducted on an automated system using more and more satellites. In particular, earth observation using satellite images is being used in various fields such as disaster prediction, damage analysis, and land cover classification. There are three main types of satellite imagery used in automation systems: optical, Synthetic Aperture Radar (SAR), and infrared. Unlike optical satellite, which is heavily influenced by weather and light, SAR satellite can acquire images in all-weather conditions. Thanks to this advantage, SAR satellite images are used in many fields, in particular, water segmentation. There are various traditional SAR image-based water segmentation methods based on thresholding technique. However, these methods are not suitable for rapidly processing a large amount of SAR images because they require a manual operation to set different thresholds for each image. In this paper, we create a large-scale dataset for water segmentation of KOrean Multi-Purpose SATellite (KOMPSAT-5) containing more than 3,000 images. We perform water segmentation using representative deep learning-based segmentation models such as Fully Convolutional Networks (FCN), U-Net, DeepUNet, and High Resolution Network (HRNet). Experimental results show that high performance of water segmentation can be obtained when a large number of training images are used for all five segmentation models. In addition, we confirm the possibility of the automatic water segmentation system from a large amount of SAR images, away from traditional manual work.

We delivered a large Introduction to Data Science course with a team of undergraduate Teaching Assistant-Researchers (TARs) who both helped students in the lab and collected qualitative observations about student learning. The TARs were concurrently participating in a senior-level Pedagogy of Data Science seminar.

Moving target indication (MTI) is considered one of the most important applications of synthetic aperture radar (SAR) in military operations and traffic monitoring. Although many studies have focused on detectors based on the statistical clutter models, a mismatch between the measured data and the applied statistical model often leads to unreliable MTI results, particularly in terms of false alarm rates. To reduce the number of false alarms, we propose an efficient MTI scheme consisting of conventional MTI techniques (the displaced phase center antenna (DPCA) and the interferogram's magnitude and phase (IMP) methods), which operate in dual-receive antenna (DRA) mode in an SAR system. These techniques are coupled with a new discrimination stage based on a new detection metric-weighted kurtosis. Using simulated and real measured data from TerraSAR-X, the proposed MTI scheme demonstrates robust performance in the reduction of false alarm rates with only a slight increase in the computation time compared with conventional schemes.

KOMPSAT (KOrea Multi-Purpose SATellite)-3A is a high-resolution earth observation satellite equipped with an electrooptic camera and infrared sensors developed by KARI. In this paper, we introduce a method of detecting ships and estimate the performance with pan-sharpened optical images of 0.55m resolution from KOMPSAT-3A. Machine learning based detection technology is essential to detect ship objects with inconsistent shapes and backgrounds, particularly in satellite images where the radiometric characteristic varies severely due to environmental factors. To cope with these inconsistencies, we introduce a ship detection algorithm using binary features and AdaBoost classification. The experimental result shows that the algorithm effectively captures the characteristics of the appearance and shows competitive detection performance.

This paper proposes a useful scheme to augment the amount of training database (DB) for ship detection using Korea multi-purpose satellite-5 (KOMPSAT-5) images and deep learning. The proposed scheme utilizes electromagnetic numerical analysis technique to generate SAR chips of ship targets. In addition, two sea clutter models are adopted to simulate realistic SAR patches containing various SAR chips. Then, the simulated SAR patches are directly used to train single shot multi-box detector (SSD) model, leading to the improvement of ship detection performance.

The Fifth KOrea Multi-Purpose SATellite (KOMPSAT-5) is the first X-band (9.66 GHz) Synthetic Aperture Radar (SAR) mission of Korea that has been operational since its launch on August 22, 2013. It has been administered and managed by the Korea Aerospace Research Institute (KARI) from the initial design, development and building to calibration/validation and the subsequent normal operation. Primary aims of the KOMPSAT-5 mission are to extend KARI's existing capability of Earth observation via optical satellites to the all-day and all-weather conditions, and to meet a range of advanced remote sensing needs in general Geographical Information System (GIS) survey and monitoring the ocean, land, ice/glacier, disaster and environment. Besides the primary aims, the KOMPSAT-5 images have been distributed to international organizations such as the International Charter since 2011 and AOGEO (Asia-Oceania Group on Earth Observation) since 2019. The KOMPSAT-5, is still operating normally beyond the original design mission operation. Imaging modes had been enhanced and added to existing modes. The system is also operating orbit maintenance for the InSAR application and the output sigma naught ( σ0) have been used for multitemporal SAR images analysis. This paper introduces overall operation, acquisition, utilization and application related to the status of the KOMPSAT-5 SAR mission. We also discuss the ways to facilitate a wider and active adoption of the KOMPSAT-5 and introduce the future continuation mission, KOMPSAT-6, that KARI is developing as her second SAR mission.

Over the past century, as display evolved, people have demanded more realistic and immersive experiences in theaters. Here, we present a tomographic projector for a volumetric display system that accommodates large audiences while providing a uniform experience. The tomographic projector combines high-speed digital micromirror and three spatial light modulators to refresh projection images at 7200 Hz. With synchronization of the tomographic projector and wearable focus-tunable eyepieces, the presented system can reconstruct 60 focal planes for volumetric representation right in front of audiences. We demonstrate proof of concept of the proposed system by implementing a miniaturized theater environment. Experimentally, we show that this system has wide expressible depth range with focus cues from 25 cm to optical infinity with sufficient tolerance while preserving high resolution and contrast. We also confirm that the proposed system provides uniform experience in a wide range of viewing zone through simulation and experiment. Additionally, the tomographic projector has capability to equalize vergence state that varies in conventional stereoscopic 3D theater according to viewing position as well as interpupillary distance. This study is concluded with thorough discussion about tomographic projectors in terms of challenges and research issues.

To perform a home service task through cooperation with a human in a real environment, a robot needs to deal with the environmental changes and accordingly plan appropriate behavior sequence. For this purpose, in this paper, we propose an adaptive task planner which is based on memory and reasoning. A robot perceives user behaviors and objects using an RGB-depth and thermal sensor. The robot stores a temporal sequence of behaviors for performing a task in its episodic memory that is realized by a sequence to sequence network. When the user command is given, the episodic memory is used to retrieve the behavior sequence to carry out the command. On the other hand, when the robot perceives user behaviors, the robot postpones its behavior till his/her behavior is stopped. Once stopped, the episodic memory retrieves the behavior sequence to conduct a task that the user has intended. A task scheduler schedules the behavior sequence from the memory and sends it to an internal simulator. The internal simulator confirms the behavior sequence to be executable and then if executable, it sends the next executable behavior to the execution module. If a behavior fails in the internal simulation test, fast forward planner generates an alternative behavior sequence to resolve the failed behavior problem. The effectiveness and applicability of the proposed planner is demonstrated by a wheel-based humanoid robot.

In the case of a stepped-frequency waveform (SFW) inverse synthetic aperture radar (ISAR) system, the translational motion (TM) of a target can be usually divided into two parts: 1) target motion within a pulse repetition interval, called the inter-pulse translational motion (IPTM) and 2) target motion between bursts, called the inter-burst translational motion (IBTM). The former induces severe blurring in the ISAR images as well as range-compressed data (i.e., range profile), and the latter also causes dramatic degradation of the ISAR image quality. In this paper, a novel framework for high-resolution gapped SFW (GSFW) ISAR imaging of high-speed maneuvering target is proposed. The main novelty of the proposed method is twofold: 1) accurate TM parameter estimation in conjunction with a compressive sensing theory using a newly devised cost function and particle swarm optimization and 2) compensation for both the IPTM and IBTM phase errors simultaneously even with the GSFW data set. Simulation results using ideal point scatterers show that the proposed method is capable of precise reconstruction of ISAR image and accurate TM parameter estimation. Experimental results using real measured data verify the robustness and the effectiveness of the proposed method.

We introduce an augmented reality near-eye display dubbed "Retinal 3D." Key features of the proposed display system are as follows: Focus cues are provided by generating the pupil-tracked light field that can be directly projected onto the retina. Generated focus cues are valid over a large depth range since laser beams are shaped for a large depth of field (DOF). Pupil-tracked light field generation significantly reduces the needed information/computation load. Also, it provides "dynamic eye-box" which can be a break-through that overcome the drawbacks of retinal projection-type displays. For implementation, we utilized a holographic optical element (HOE) as an image combiner, which allowed high transparency with a thin structure. Compared with current augmented reality displays, the proposed system shows competitive performances of a large field of view (FOV), high transparency, high contrast, high resolution, as well as focus cues in a large depth range. Two prototypes are presented along with experimental results and assessments. Analysis on the DOF of light rays and validity of focus cue generation are presented as well. Combination of pupil tracking and advanced near-eye display technique opens new possibilities of the future augmented reality.

To prepare for the anticipated age of human-robot symbiosis, robots should be able to interact and cooperate with humans effectively by understanding the meaning and intention of human behavior. In this paper, we define human intention as “desired behavior of the human using objects.” To infer the defined human intention, a robot should learn the object affordance along with a behavior hierarchy structure. Thus, in this paper, we propose a behavior hierarchy-based affordance network (BHAN) and a behavior hierarchy-based affordance map (BHAM) to represent the object affordance, behavior hierarchy structure, and object hierarchy structure, simultaneously. Autonomous and interactive BHAN/BHAM learning algorithms are also proposed to make a robot develop the BHAN and BHAM by itself, as well as by interacting with a human. Based on the newly developed BHANs and BHAM, a robot could infer the human intention from information observed in context and from human behavior. The effectiveness of the proposed method was demonstrated through experiments on human-robot interaction with building blocks using a simulated differential wheel robot and a real human-sized humanoid robot.

We propose a see-through additive light field display as a novel type of compressive light field display. We utilize holographic optical elements (HOEs) as transparent additive layers. The HOE layers are almost free from diffraction unlike spatial light modulator layers, which makes this additive light field display more advantageous when modifying the number of layers, thickness, and pixel density compared with conventional compressive displays. Meanwhile, the additive light field display maintains advantages of compressive light field displays. The proposed additive light field display shows bright and full-color volumetric images in high definition. In addition, users can view real-world scenes beyond the displays. Hence, we expect that our method can contribute to the realization of augmented reality. Here, we describe implementation of a prototype additive light field display with two additive layers, evaluate the performance of transparent HOE layers, describe several results of display experiments, discuss the diffraction effect of spatial light modulators, and analyze the ability of the additive light field display to express uncorrelated light fields.

This paper describes a miniature, wearable wireless hand- gesture recognition (HGR) system called HANDIO. It obtains its input from not only traditional inertial sensors but also a muscle tension sensor (MTS). The addition of MTS enables recognition of a much broader range of intuitive hand gestures, particularly those involving the wrist, that would otherwise be difficult to distinguish by traditional inertial-only HGRs. Among MTSs, we choose an optical MTS over the conventional surface electromyography(sEMG) for the small size, low power consumption, wearing comfort, and good detection rate. This novel miniaturized design enables the whole system to be easily patched on the wrist area or integrated into a wearable device such as a wristband or watch without extra wiring. Experimental results show that a total of 8 hand gestures involving the wrist can be recognized with a detection rate over 93%. The average power consumption of the optical sensor is only around 258μW. This versatile system can also be used to detect other joint activities such as the elbow and knee joint.

Sternocleidomastoid (SCM) is a paired muscle that stretches along both sides of the neck area. It acts as an accessory muscle of inhalation. Abnormal SCM contraction during asthma is usually a sign of further respiratory impairment. Thus, monitoring SCM muscles has great significance in asthma assessment and control. In this work, we develop a wearable monitoring system based on an optical sensor that consists of an LED and a photo detector (PD). A voltage comparator enables the microcontroller unit (MCU) to remain in sleep mode until waken upon detecting contraction. Experimental results show that our optical sensor consumes much lower power than surface electromyography (sEMG), the most commonly used technique while offering more comfort and compactness. It is also robust to motion artifact and DC baseline wandering. These properties simplify the hardware design, while the use of the comparator further reduces the system power consumption to >450 μW on average, making it the best option for low power monitoring.

This work investigates the theoretical analysis to enable fast and accurate estimation of dominant frequencies from randomly sampled signals by compressive sensing (CS). We show that dominant frequencies can be discovered using partially computed Discrete Cosine Transform (DCT). We also propose a new system structure with an estimation unit that enables the signal reconstruction to be selectively bypassed for CS-based devices on signals with dominant frequencies, thus increasing the responsiveness and further reducing the power consumption. For verification, we design a photoplethysmagram (PPG) based heart rate monitor using the proposed algorithm. The accuracy is tested using MIMIC database. The detected heart rate is within 1 beat per minute from the reference over 99% of the data.

In this paper, we compared the performance of bistatic radar target identification using the computed bistatic RCS of the full-scale targets. The FFT (fast Fourier transform)-based CLEAN is used as the feature vector extraction method and multi-layered perceptron (MLP) neural network is used as a classifier. Simulation results show that the optimally positioned bistatic radar has better target identification performance, demonstrating the importance of the transmitter and receiver positions in bistatic radar.

The piracy of copyrighted digital content over the Internet infringes copyrights and damages the digital content industry. Accordingly, identifying and monitoring technology on the online content service like fingerprinting is getting valuable through the explosion of digital content sharing. This paper proposes a robust video fingerprinting feature to identify a modified video clip from a large scale database. Hierarchical symmetric difference feature is proposed in order to offer efficient video fingerprinting. The feature is robust and pairwise independent against various video modifications such as compression, resizing, or cropping. Moreover, videos undergoing a transformation such as flipping or mirroring can be identified by simply disordering the bit pattern of fingerprints. The performance of the proposed feature is extensively experimented on 6,482 hours of database and the experimental results show that the proposed fingerprinting is efficient and robust against various modifications.

In this paper, we present a mood-based music recommendation player: Smoodi. The Smoodi provides smart mood recommendation by touch and drag and it has three different views: Mood Square, Cover Flow and Mood Cloud. In the Mood Square, users can check the mood distribution of local clips and generate playlists by touching mood cells. In the Cover Flow, users can find music information and generate similar songs list by dragging a seed song. In the Tag Cloud, users can make playlists by selecting mood tags. For this application, we developed new mood model from collected mood tags and arousal-valence position values and designed a regression function to estimate mood probabilities.

The authors present an approach that extracts human activity patterns of daily living and represents spatiotemporal relations between activities intuitively. In general, customised services are provided based on activity patterns of users. This study focuses on extracting and determining activities that occur simultaneously. In order to determine simultaneous activities, the authors analysed the daily activities that are collected from device applications such as location sensors and electronics. In addition, a context model using the incremental statistical method is organised and temporal relations between the activities patterns are analysed. Furthermore, information visualisation of the spatiotemporal topology with duration and frequency is demonstrated. Also, the authors have experimented on a test-bed called the ubiquitous smart space and compared the accuracy of the incremental statistical method with that of the non-incremental method.

In this paper, we propose a video fingerprinting method based on the orientation of luminance centroid. To attain robustness against frame rate change, variation in color characteristics and resizing, video frames are normalized in time, spatial and color domain. Thereafter, the orientation of luminance centroid is calculated as as video fingerprint. Under various distortions, the proposed method is evaluated and compared with other centroid-based methods. Experimental results show that the proposed method satisfies the requirements of video fingerprint, and performance comparison with other methods in terms of receiver operating characteristic indicates that the proposed method is more robust and has better discriminatory property than others.

Most sensor networks currently used to gather sensing data from a broad environment in which it is very difficult to deploy the existing networks (e.g., internet). Recently, researches on relational database approaches to sensor networks are being tried. There occur, however, some problems in applying directly the traditional relational database concepts into a sensor network. One reason is because the join operator only allows to perform operations on tuples which has exactly same join attribute values. For instance, let us assume a sensor network that two different classes of nodes are randomly scattered in the same area. We cannot get join results to know the relationship between two different classes of sensing values because there might be no nodes at a exactly same location. For a solution about the above described problem we propose in this paper new join operators. These new join operators can provide more effective data management and standard interfaces to application programs in sensor networks.

Recently, researches on relational database approaches to sensor networks are being tried. There occur, however, some problems in applying directly the traditional relational database concepts into a sensor network in that every database operation is performed only on the real existing data which are tuples in database relations. The reason is because in a sensor network viewpoint situations under which some operations should be performed on non-existing data may occur frequently. For instance, Can we write a query to get temperature of a spot with no sensor node? Additionally, let us assume a sensor network that two different classes of nodes are scattered in the same area. We cannot get join results to know the relationship between two different classes of sensing values because there might be no nodes at a exact same location. For a solution about the above described problems we propose in this paper new database operators. This new database operators can provide more effective data management and standard interfaces to application programs in sensor networks.

In this paper, we propose a wavelet-based audio watermarking algorithm that is robust against capturing attack. With a commercial capturing tool, it is possible to capture various audio contents without noticeable degradation, and thus can potentially facilitate the illegal distribution of the audio content. By adjusting the mean value of the lowest subband coefficients of the discrete wavelet transform (DWT) of the audio, the proposed watermark can survive capturing attack including sampling rate conversion, random cropping and compression. By incorporating a simple human auditory model, the inaudibility of the watermark achieved, and the detection probability is improved based on the difference information of extracted values. This is confirmed by experimental results

The performance of a fingerprinting system, which is often measured in terms of reliability and robustness, is directly related to the features that the system uses. In this letter, we present a new audio-fingerprinting method based on the normalized spectral subband moments. A threshold used to reliably determine a fingerprint match is obtained by modeling the features as a stationary process. The robustness of the normalized moments was evaluated experimentally and compared with that of the spectral flatness measure. Among the considered subband features, the first-order normalized moment showed the best performance for fingerprinting.

For image content authentication, a secure watermarking method using quantization-based embedding on the largest singular value (SV) is proposed. The block-wise quantization-based embedding can be vulnerable to vector quantization (VQ) attack and attacks associated with histogram analysis. To overcome these security problems, the proposed method places interdependency among image blocks and dithers the quantized value. By adjusting the threshold of the detector, a trade-off between the robustness to JPEG compression and the probability of misdetection can be made. The proposed method can detect a tampered area with high sensitivity. This is confirmed by experimental results and security analysis.

For multimedia fingerprinting, it is crucial to extract relevant features that allow direct access to the distinguishing characteristics of a multimedia object. Features used for fingerprinting directly relate to the performance of the entire fingerprinting system. The paper proposes a novel audio fingerprinting method based on normalized spectral subband centroids. The spectral subband centroid is selected due to its resilience against equalization, compression, and noise addition. Both reliability and robustness issues in the fingerprinting system are addressed. Experimental results show that the proposed method is not only reliable, but also robust against various audio processing steps, including MP3 compression, equalization, random start, time-scale modification, and linear speed change.