Searcharxiv⌕ Search

arXiv subjects

Mohamed Younis

Publications and source records attributed to Mohamed Younis.

At least 19 recordsLinked to original sources

Arbor: When Labels Are Not Coarsened at Random

Merging datasets labelled at different levels of a taxonomy leaves some images with only a coarse label, and some classes with no fine-grained label at all. Standard training scores a coarse label by the total probability of the classes under it. Coarse labels name superclasses, and the standard objective is exact only when every class within a superclass is equally likely to receive a coarse label. We call a class novel if it is never labelled finely, and known otherwise. A novel class violates the assumption, and the classifier assigns its images to a known sibling. We propose Arbor, which models how labels become coarse. In each superclass with a novel class, Arbor adds a virtual novel class, and a logistic gate divides the superclass's probability between it and the known classes. The gate is trained with the likelihood of the observed labels under this model, the coarsening likelihood. We prove that at its population optimum the gate equals the Elkan-Noto correction of positive-unlabelled learning, and show that training the correction into the classifier improves on applying it after training. Across Cars, CUB and Aircraft with 90% coarse labels, Arbor correctly assigns 76-85% of novel-class images to their superclass, more than hierarchical classifiers, hierarchical novelty detection and the Elkan-Noto correction. Its top-1 accuracy over the leaf classes on known-class images is within 1.6 points of the best baseline, although its final predictions return the superclass more often. On the more challenging iNaturalist-2019 benchmark, Arbor improves novel-class accuracy from 46% (uniform target, the strongest baseline) to 68%.

cs.CV↗

SEALECT: Scene-Conditioned Pose Priors for Plausible Maritime Vessel Retrieval

We address scene-conditioned object retrieval for compositional augmentation in maritime imagery, where strong geometric constraints make appearance-only retrieval produce physically implausible vessel insertions. We propose SEALECT, a retrieval framework that jointly models visual compatibility, vessel pose, scale, and scene geometry through a unified score combining neural similarity, scene-conditioned pose priors, and observed-pose consistency. We construct a maritime scene-pose library and model the pose and scale priors nonparametrically, as a kernel that conditions each prior continuously on the scene with no learned parameters. Across seven maritime benchmarks, our trained scene-conditioned retriever exceeds the strongest appearance encoder (SigLIP 2) in recall (+0.14), and the training-free prior raises geometric plausibility further at negligible recall cost, more than halving the constraint violations of appearance retrieval and outperforming learned foreground-search baselines (GALA-style, DiscoFOS). In a blind study of end-to-end composites (BootPlace placement, DiffHarmony harmonization), SEALECT ranks first among six methods, ahead of random and GALA-style selection and statistically tied with CLIP, DiscoFOS, and DINO.

cs.CV↗

HullProbe: Measuring Interventional Part Reliance in Maritime Vessel Classifiers

Recognizing a vessel's type from an image supports maritime surveillance and sanctions monitoring, but a classifier can reach high accuracy by leaning on the generic hull instead of the more distinctive superstructure, and is then fooled when a vessel is re-hulled or occluded. De-biasing methods claim to remove such reliance, yet the field certifies them with worst-group accuracy, a proxy that needs group labels and only infers reliance from accuracy. We measure the reliance directly. HullProbe swaps a registered anatomical part of a vessel for another instance's and reads how much that part drives the decision, using a same-class swap to control for compositing artifacts. Because a surface vessel is a hull and a superstructure arranged along one axis, its parts form a 2x3 grid whose 2^6 = 64 swaps we evaluate exhaustively, and a single Walsh transform returns each part's interventional influence and its interaction spectrum. On planted shortcuts, where the true reliance is known, the measure tracks worst-group accuracy (r = -0.94 over 45 configurations), and replacing its learned segmenter with human annotations changes the hull-share by less than the audit's resolution. On the 26 vessel types of MARVEL, the classifier places about 46% of its corrected sensitivity on the generic hull yet follows the superstructure when the two conflict, and the group-label-free de-biasers available there do not detectably reduce the hull's share. The findings hold across three architectures and a second maritime dataset, and the method extends to fine-grained vehicle recognition on Stanford Cars.

cs.CV↗

Shift-Reliance Alignment: Ranking Vision-Language Models by the Cues They Trust

Vision-language models (VLMs) are usually chosen from leaderboards, which give one ranking for every application. On questions about objects that a scene suggests but does not contain, however, the model with the lowest error on the POPE hallucination benchmark is not the one with the lowest error on such questions. For yes/no questions about object presence, we show that the best model for a target domain can be predicted without running any candidate on that domain, given 650 to 1,200 labeled target items that calibrate a detector. We measure each model's reliance on a cue, a feature such as scene context or text in the image, by editing 100 labeled source images so that the cue suggests the wrong answer, and comparing with a matched control edit. Reliance predicts which VLMs are misled by physical sticker attacks (Spearman .93 on nine models not used to design the score) and which hallucinate on unedited COCO photographs, beyond what their tendency to answer yes explains. Shift-Reliance Alignment (SRA) scores a model by its source error on the labeled source questions plus its reliance multiplied by the shift. The shift is the proportion of target questions on which the cue misleads, and a detector estimates it on the unlabeled target images. SRA selects the model with the lowest score. In pre-registered tests on new images where most questions mislead, it lowered the error of the selected model from 30.4% under selection by source error to 18.8% on Objects365 and OpenImages but, against its prediction, not on COCO. On CueDial, a benchmark we release that varies how often a cue misleads, selection by model confidence falls more than 40 percentage points behind the best model when overlaid text denies a visible object, and SRA stays within 3 points.

cs.CV↗

SWORD: A Secure LoW-Latency Offline-First Authentication and Data Sharing Scheme for Resource Constrained Distributed Networks

While many resource-constrained networks, such as Internet of Things (IoT) and Internet of Vehicles (IoV), are inherently distributed, the majority still rely on central servers for fast authentication and data sharing. Blockchain-based solutions offer decentralized alternatives but often struggle to meet the stringent latency requirements of real-time applications. Even with the rollout of 5G, network latency between servers and peers remains a significant challenge. To address this, we introduce SWORD, a novel offline-first authentication and data-sharing scheme designed specifically for resource-constrained networks. SWORD utilizes a proximity-based clustering approach to enable offline authentication and data sharing, ensuring low-latency, secure operations even in intermittently connected scenarios. Our experimental results show that SWORD outperforms traditional blockchain-based solutions while offering similar resource efficiency and authentication latency to central-server-based solutions. Additionally, we provide a comprehensive security analysis, demonstrating that SWORD is resilient against spoofing, impersonation, replay, and man-in-the-middle attacks.

cs.CR↗

A novel pedestrian road crossing simulator for dynamic traffic light scheduling systems

The major advances in intelligent transportation systems are pushing societal services toward autonomy where road management is to be more agile in order to cope with changes and continue to yield optimal performance. However, the pedestrian experience is not sufficiently considered. Particularly, signalized intersections are expected to be popular if not dominant in urban settings where pedestrian density is high. This paper presents the design of a novel environment for simulating human motion on signalized crosswalks at a fine-grained level. Such a simulation not only captures typical behavior, but also handles cases where large pedestrian groups cross from both directions. The proposed simulator is instrumental for optimized road configuration management where the pedestrians' quality of experience, for example, waiting time, is factored in. The validation results using field data show that an accuracy of 98.37 percent can be obtained for the estimated crossing time. Other results using synthetic data show that our simulator enables optimized traffic light scheduling that diminishes pedestrians' waiting time without sacrificing vehicular throughput.

eess.SY↗

Dynamic Road Management in the Era of CAV

Traffic management and on-road safety have been a concern for the transportation authorities and the engineering communities for many years. Most of the implemented technologies for intelligent highways focus on safety measures and increased driver awareness, and expect a centralized management for the vehicular traffic flow. Leveraging recent advances in wireless communication, researchers have proposed solutions based on vehicle-to-vehicle (V2V) and vehicle-to-Infrastructure (V2I) communication in order to detect traffic jams and better disseminate data from on-road and on-vehicle sensors. Moreover, the development of connected autonomous vehicles (CAV) have motivated a paradigm shift in how traffic will be managed. Overall, these major technological advances have motivated the notion of dynamic traffic management (DTM), where smart road reconfiguration capabilities, e.g., dynamic lane reversal, adaptive traffic light timing, etc. will be exploited in real-time to improve traffic flow and adapt to unexpected incidents. This chapter discusses what the challenges in realizing DTM are and covers how CAV has revolutionized traffic management. Moreover, we highlight the issues for handling human-driven vehicles while roads are transitioning to CAV only traffic. Particularly, we articulate a new vision for inter-vehicle communication and assessment of road conditions, and promote a novel system for traffic management. Vehicle to on-road sensors as well as inter-vehicle connectivity will be enabled through the use of handheld devices such as smartphones. This not only enables real-time data sharing but also expedites the adoption of DTM without awaiting the dominant presence of autonomous vehicle on the road. ...

cs.NI↗

Vapor Cloud Delayed-DPPM Modulation Technique for nonlinear Optoacoustic Communication

The optoacoustic process can solve the longstanding challenge of wireless information transmission from an airborne unit to an underwater node (UWN). The nonlinear optoacoustic signal generated by proper laser parameters can propagate long distances in water. However, forming such a signal requires a high-power laser, and the buildup of a vapor cloud precludes the subsequent acoustic signal generation. Therefore, pursuing the traditional on-off keying (OOK) modulation technique will limit the data rate and power efficiency. In this paper, we analyze different modulation techniques and propose a vapor cloud delayed-differential pulse position modulation (VCD-DPPM) technique to improve the data rate and achieve high power efficiency for a single stationary laser transmitter. The symbol rate of VCD-DPPM is approximately 6.9 times and 1.69 times higher than OOK in our text communication simulation using a laser repetition rate of 10 kHz and 40 Hz, respectively. Furthermore, VCD-DPPM is 137% more power efficient than the OOK technique for both cases. We have generated different acoustic signal levels in laboratory conditions and simulated the bit error rate (BER) for different depths and positions of the UWN, while considering ambient underwater noises. Our results indicate that VCD-DPPM enables efficient data transmission.

eess.SP↗

An Attack Resilient PUF-based Authentication Mechanism for Distributed Systems

In most PUF-based authentication schemes, a central server is usually engaged to verify the response of the device's PUF to challenge bit-streams. However, the server availability may be intermittent in practice. To tackle such an issue, this paper proposes a new protocol for supporting distributed authentication while avoiding vulnerability to information leakage where CRPs could be retrieved from hacked devices and collectively used to model the PUF. The main idea is to provision for scrambling the challenge bit-stream in a way that is dependent on the verifier. The scrambling pattern varies per authentication round for each device and independently across devices. In essence, the scrambling function becomes node- and packet-specific and the response received by two verifiers of one device for the same challenge bit-stream could vary. Thus, neither the scrambling function can be reverted, nor the PUF can be modeled even by a collusive set of malicious nodes. The validation results using data of an FPGA-based implementation demonstrate the effectiveness of our approach in thwarting PUF modeling attacks by collusive actors. We also discuss the approach resiliency against impersonation, Sybil, and reverse engineering attacks.

cs.CR↗

Underwater Node Localization using Optoacoustic Signals

Localization of underwater networks is important in many military and civil applications. Because GPS receivers do not work below the water surface, traditional localization methods form a relative topology of underwater nodes (UWNs) and utilize either anchor nodes or floating gateways with dual transceivers in order to determine global coordinates. However, these methods introduce logistical complications and security risks in deploying the anchor and/or surface gateways. This paper tackles such an issue by proposing new localization techniques which can remotely localize UWNs using optoacoustic signals. In our approach, GPS coordinates are transmitted from air to the UWN via creating an underwater temporary isotropic acoustic transmitter with the optoacoustic process. We analyze the process of controlling the shape and size of the plasma to create the isotropic acoustic transmitter and experimentally validate the generation of isotropic acoustic signals. Then two methods of localization are proposed for static and dynamic UWNs. Finally, the simulation results with experimental values show the effectiveness of our approach. Comparing to the traditional techniques, our approach achieves the same accuracy without using any surface or underwater anchor nodes.

eess.SP↗

From MANET to people-centric networking: milestones and open research challenges

In this paper we discuss the state of the art of (mobile) multi-hop ad hoc networking with the aim to present the current status of the research activities and identify the consolidated research areas, with limited research opportunities, and the hot and emerging research areas for which further research is required. We start by briefly discussing the MANET paradigm, and why the research on MANET protocols is now a cold research topic. Then we analyze the active research areas. Specifically, after discussing the wireless-network technologies we analyze four successful ad hoc networking paradigms, mesh, opportunistic, vehicular networks, and sensor networks that emerged from the MANET world. We also present the new research directions in the multi-hop ad hoc networking field: people-centric networking, triggered by the increasing penetration of the smartphones in everyday life, which is generating a people-centric revolution in computing and communications.

cs.NI↗

Green-PoW: An Energy-Efficient Blockchain Proof-of-Work Consensus Algorithm

This paper opts to mitigate the energy-inefficiency of the Blockchain Proof-of-Work (PoW) consensus algorithm by rationally repurposing the power spent during the mining process. The original PoW mining scheme is designed to consider one block at a time and assign a reward to the first place winner of a computation race. To reduce the mining-related energy consumption, we propose to compensate the computation effort of the runner(s)-up of a mining round, by granting them exclusivity of solving the upcoming block in the next round. This will considerably reduce the number of competing nodes in the next round and consequently, the consumed energy. Our proposed scheme divides time into epochs, where each comprises two mining rounds; in the first one, all network nodes can participate in the mining process, whereas in the second round only runners-up can take part. Thus, the overall mining energy consumption can be reduced to nearly $50\%$. To the best of our knowledge, our proposed scheme is the first to considerably improve the energy consumption of the original PoW algorithm. Our analysis demonstrates the effectiveness of our scheme in reducing energy consumption, the probability of fork occurrences, the level of mining centralization presented in the original PoW algorithm, and the effect of transaction censorship attack.

cs.CR↗

Enabling efficient application monitoring in cloud data centers using SDN

Software Defined Networking (SDN) not only enables agility through the realization of part of the network functionality in software but also facilitates offering advanced features at the network layer. Hence, SDN can support a wide range of middleware services; network performance monitoring is an example of these services that are already deployed in practice. In this paper, we exploit the use of SDNs to efficiently provide application monitoring functionality. The recent rise of complex cloud applications has made performance monitoring a major issue. We show that many performance indicators can be inferred from messages exchanged among application components. By analyzing these messages, we argue that the overhead of performance monitoring could be effectively moved from the end hosts into the SDN middleware of the cloud infrastructure which enables more flexible placement of logging functionality. This paper explores several approaches for supporting application monitoring through SDN. In particular, we combine selective forwarding in SDN to enable message filtering and reformatting, and propose a customized port sniffing technique. We describe the implementation of the approach within the standard SDN software, namely OVS. We further provide a comprehensive performance evaluation to analyze advantages and disadvantages of our approach, and highlight the trade-offs.

cs.NI↗

IoT-enabled Channel Selection Approach for WBANs

-Recent advances in microelectronics have enabled the realization of Wireless Body Area Networks (WBANs). However , the massive growth in wireless devices and the push for interconnecting these devices to form an Internet of Things (IoT) can be challenging for WBANs; hence robust communication is necessary through careful medium access arbitration. In this paper, we propose a new protocol to enable WBAN operation within an IoT. Basically, we leverage the emerging Bluetooth Low Energy technology (BLE) and promote the integration of a BLE transceiver and a Cognitive Radio module (CR) within the WBAN coordinator. Accordingly, a BLE informs WBANs through announcements about the frequency channels that are being used in their vicinity. To mitigate interference, the superframe's active period is extended to involve not only a Time Division Multiple Access (TDMA) frame, but also a Flexible Channel Selection (FCS) and a Flexible Backup TDMA (FBTDMA) frames. The WBAN sensors that experience interference on the default channel within the TDMA frame will eventually switch to another Interference Mitigation Channel (IMC). With the help of CR, an IMC is selected for a WBAN and each interfering sensor will be allocated a time-slot within the (FBTDMA) frame to retransmit using such IMC.

cs.NI↗

Efficient Medium Access Arbitration Among Interfering WBANs Using Latin Rectangles

The overlap of transmission ranges among multiple Wireless Body Area Networks (WBANs) is referred to as coexistence. The interference is most likely to affect the communication links and degrade the performance when sensors of different WBANs simultaneously transmit using the same channel. In this paper, we propose a distributed approach that adapts to the size of the network, i.e., the number of coexisting WBANs, and to the density of sensors forming each individual WBAN in order to minimize the impact of co-channel interference through dynamic channel hopping based on Latin rectangles. Furthermore, the proposed approach opts to reduce the overhead resulting from channel hopping, and lowers the transmission delay, and saves the power resource at both sensor- and WBAN-levels. Specifically, we propose two schemes for channel allocation and medium access scheduling to diminish the probability of inter-WBAN interference. The first scheme, namely, Distributed Interference Avoidance using Latin rectangles (DAIL), assigns channel and time-slot combination that reduces the probability of medium access collision. DAIL suits crowded areas, e.g., high density of coexisting WBANs, and involves overhead due to frequent channel hopping at the WBAN coordinator and sensors. The second scheme, namely, CHIM, takes advantage of the relatively lower density of collocated WBANs to save power by hopping among channels only when interference is detected at the level of the individual nodes. We present an analytical model that derives the collision probability and network throughput. The performance of DAIL and CHIM is further validated through simulations.

cs.NI↗

Distributed Scheme for Interference Mitigation of WBANs Using Predictable Channel Hopping

When sensors of different coexisting wireless body area networks (WBANs) transmit at the same time using the same channel, a co-channel interference is experienced and hence the performance of the involved WBANs may be degraded. In this paper, we exploit the 16 channels available in the 2.4 GHz international band of ZIGBEE, and propose a distributed scheme that avoids interference through predictable channel hopping based on Latin rectangles, namely, CHIM. In the proposed CHIM scheme, each WBAN's coordinator picks a Latin rectangle whose rows are ZIGBEE channels and columns are sensor IDs. Based on the Latin rectangle of the individual WBAN, each sensor is allocated a backup time-slot and a channel to use if it experiences interference such that collisions among different transmissions of coexisting WBANs are minimized. We further present a mathematical analysis that derives the collision probability of each sensor's transmission in the network. In addition, the efficiency of CHIM in terms of transmission delay and energy consumption minimization are validated by simulations.

cs.NI↗

Distributed Scheme for Interference Mitigation of Coexisting WBANs Using Latin Rectangles

The performance of wireless body area networks (WBANs) may be degraded due to co-channel interference, i.e., when sensors of different coexisting WBANs transmit at the same time-slots using the same channel. In this paper, we exploit the 16 channels available in the 2.4 GHz unlicensed international band of ZIGBEE, and propose a distributed scheme that opts to avoid interference through channel to time-slot hopping based on Latin rectangles, DAIL. In DAIL, each WBAN's coordinator picks a Latin rectangle whose rows are ZIGBEE channels and columns are time-slots of its superframe. Subsequently, it assigns a unique symbol to each sensor; this latter forms a transmission pattern according to distinct positions of its symbol in the rectangle, such that collisions among different transmissions of coexisting WBANs are minimized. We further present an analytical model that derives bounds on the collision probability of each sensor's transmission in the network. In addition, the efficiency of DAIL in interference mitigation has been validated by simulations.

cs.NI↗

Inter-WBANs Interference Mitigation Using Orthogonal Walsh Hadamard Codes

A Wireless Body Area Network (WBAN) provides health care services. The performance and utility of WBANs can be degraded due to interference. In this paper, our contribution for co-channel interference mitigation among coexisting WBANs is threefold. First, we propose a distributed orthogonal code allocation scheme, namely, OCAIM, where, each WBAN generates sensor interference lists (SILs), and then all sensors belonging to these lists are allocated orthogonal codes. Secondly, we propose a distributed time reference correlation scheme, namely, DTRC, that is used as a building block of OCAIM. DTRC enables eachWBAN to generate a virtual time-based pattern to relate the different superframes. Accordingly, DTRC provides each WBAN with the knowledge about, 1) which superframes and, 2) which time-slots of those superframes interfere with the time-slots within its superframe. Thirdly, we further analyze the success and collision probabilities of frames transmissions when the number of coexisting WBANs grows. The simulation results demonstrate that OCAIM outperforms other competing schemes in terms of interference mitigation and power savings.

cs.NI↗