57,417 research outputs found
Non-Coherent Cooperative Communications Dispensing with Channel Estimation Relying on Erasure Insertion Aided Reed-Solomon Coded SFH M-ary FSK Subjected to Partial-Band Interference and Rayleigh Fading
No full textThe rationale of our design is that although much of the literature of cooperative systems assumes perfect coherent detection, the assumption of having any channel estimates at the relays imposes an unreasonable burden on the relay station. Hence, non-coherently detected Reed-Solomon (ReS) coded Slow Frequency Hopping (SFH) assisted M -ary Frequency Shift Keying (FSK) is proposed for cooperative wireless networks, subjected to both partial-band interference and Rayleigh fading. Erasure insertion (EI) assisted ReS decoding based on the joint maximum output-ratio threshold test (MO-RTT) is investigated in order to evaluate the attainable system performance. Compared to the conventional error-correction-only decoder, the EI scheme may achieve an Eb/N0 gain of approximately 3dB at the Codeword Error Probability, Pw , of 10-4 , when employing the ReS (31, 20) code combined with 32-FSK modulation. Additionally, we evaluated the system’s performance, when either equal gain combining (EGC) or selection combining (SC) techniques are employed at the destination’s receiver. The results demonstrated that in the presence of one and two assisting relays, the EGC scheme achieves gains of 1.5 dB and 1.0 dB at the Pw of 10-6 , respectively, compared to the SC arrangement. Furthermore, we demonstrated that for the same coding rate and packet size, the ReS (31, 20) code using EI decoding is capable of outperforming convolutional coding, when 32-FSK modulation is considered, whilst LDPC coding had an edge over the above two schemes
Unlocking robotic perception: comparison of deep learning methods for simultaneous localization and mapping and visual simultaneous localization and mapping in robot
No full textSimultaneous Localization and Mapping (SLAM) and Visual SLAM are crucial technologies in robotics, allowing autonomous systems to navigate and comprehend their environment. Deep learning (DL) has become a powerful tool in driving progress in these areas, providing solutions that improve accuracy, efficiency, and resilience. This article thoroughly analyzes different deep learning techniques designed explicitly for SLAM and Visual SLAM applications in robotic systems. This work provides a detailed overview of DL roles in SLAM and VSLAM and emphasizes the differences between these two fields. Five powerful DL methods are investigated: Convolutional Neural Networks in extracting features and understanding meaning, Recurrent Neural Network in modeling temporal relationships, Deep Reinforcement Learning in developing exploration strategies, Graph Neural Network in modeling spatial relationships, and Attention Mechanisms in selectively processing information. In this research, we will examine the advantages and disadvantages of each approach in relation to robotic applications, taking into account issues such as real-time performance, resource restrictions, and adaptability to various situations. This article seeks to guide researchers and practitioners in selecting suitable deep learning algorithms to improve the capabilities of SLAM and Visual SLAM in robotic systems by combining ideas from recent research and actual implementations. The popular types of each concerned DL will be synthesized with the discussion of pros and cons
Digital Twin-Based Real-Time Monitoring System for Safety of Multiple Laptops in Working Environment
No full textOverheating is a significant issue for laptops, especially in working environments where multiple laptops are utilized to launch heavy programs without the user's presence frequently. The proposed monitoring method is based on a digital twin (DT) system in a workstation that monitors the heat power loss of the battery, relying on the measurements of the battery current, the central processing units (CPUs) temperatures, and graphics processing units (GPUs) temperatures. Other personal computer (PC) laptops deliver their data via transmission control protocol (TCP)/Internet protocol (IP). The workstation sends a query message, allowing each PC to transfer the necessary information. The virtual system handles the received data to proceed with the alert system for warning if these parameters overcome the concerned threshold. The DT system is implemented into a Dell workstation based on MATLAB-Simulink to manage the data from ten other PCs in the experiment. Power loss is calculated in the DT Simulink system based on the electrical circuit of the PC in plugged-in mode and discharge mode. Together with other factors, the system successfully monitors and detects the current situation with a specific alert function. The monitored parameters can be observed for each setting-up period, which allows the manager to comprehend the condition state during a specific time
New Artificial Intelligence Approach to Inclination Measurement Based on MEMS Accelerometer
No full textThe article presents a research of angular orientation based on a microelectromechanical system (MEMS) accelerometer by using machine learning (ML) and deep learning (DL) model with architectures of deep neural networks (DNNs). In the industrial environment, artificial intelligence (AI) plays a crucial role in automation which is a potential solution for better performance of inclinometer. This article was carried out to apply this intelligent model on the inertial measurement unit to accomplish the angular position. The experiment shows that the ML model correctly learns the relationship between acceleration and tracking angles via polynomial regression with an R-square of 0.98. The employed DL model with four hidden layers of ten neurons achieves an accuracy of 99.99 % and almost a nonerror performance. The acceleration acquisitions were obtained from MEMS accelerometer LSM9DS1 at a frequency of 50 Hz via microcontroller STM32F401RE. The ML and DNN models were designed based on the platform Tensorflow with high processing accuracy. The Pan-Tilt Unit was used as the angle reference for static and dynamic tests. The traditional technique is used for comparison as well as verification of the proposed models. The DL model has better precision over the ML model due to its high structure level with updating weight and error optimization from the neural network structure. Meanwhile, ML shows more stable results in dynamic circumstances
Leptolalax bidoupensis Rowley, Le, Tran & Hoang, 2011, sp. nov.
No full textLeptolalax bidoupensis sp. nov. Holotype: AMS R 173133, adult male, calling on clay bank 0.2 m from 1–4 m wide, medium-high gradient, rocky stream in montane evergreen forest in Bidoup-Nui Ba National Park, Lam Dong Province, Vietnam (12.19225 º N, 108.71494 º E, 1730 m, Figure 1). Collected at 23: 55 h on 19 May 2008 by J. J. L. Rowley, Hoang D. H., Le T. T. D., and Tran T. A. D. Paratypes: UNS 00101 /AMS R 173135, adult male, calling on tree root 0.2 m above 2–5 m wide, mediumhigh gradient, rocky stream in montane evergreen forest in Bidoup-Nui Ba National Park, Lam Dong Province, Vietnam (12.19106 º N, 108.71703 º E, 1641 m), collected at 20: 25 h on 20 May 2008. UNS 00102 /AMS R 173137, metamorph, in water of swampy area adjacent to a swift, rocky stream in cloud forest in Bidoup-Nui Ba National Park, Lam Dong Province, Vietnam (12.18644 º N, 108.71486 º E, 1627 m), collected at 19: 45 h on 18 May 2008. AMS R 173134, adult female, on clay bank 0.2 m from 1–4 m wide, medium-high gradient, rocky stream in montane evergreen forest in Bidoup-Nui Ba National Park, Lam Dong Province, Vietnam (12.19225 º N, 108.71494 º E, 1730 m), collected at 23: 50 h on 19 May 2008, in close proximity to holotype. AMS R 173136, adult male, calling on leaf litter 0.2 m from 2–5 m wide, medium-high gradient, rocky stream in montane evergreen forest in Bidoup- Nui Ba National Park, Lam Dong Province, Vietnam (12.19106 º N, 108.71703 º E, 1641 m), collected at 21: 40 h on 20 May 2008. NCSM 77320, adult female, in water of swampy area off swift, rocky stream in montane evergreen forest in Bidoup-Nui Ba National Park, Lam Dong Province, Vietnam (12.18644 º N, 108.71486 º E, 1627 m), collected at 19: 50 h on 18 May 2008. NCSM 77321, adult male, sitting in upright posture (previously calling?) in leaf litter, 0.1 m from swift, rocky stream in montane evergreen forest in Bidoup-Nui Ba National Park, Lam Dong Province, Vietnam (12.18644 º N, 108.71486 º E, 1627 m), collected at 21: 45 h on 18 May 2008. NCSM 77322, metamorph, on clay bank 0.5 m from 2–5 m wide, medium-high gradient, rocky stream in montane evergreen forest in Bidoup-Nui Ba National Park, Lam Dong Province, Vietnam (12.19106 º N, 108.71703 º E, 1641 m), collected at 21: 30 h on 20 May 2008. All specimens were collected by J. J. L. Rowley, Hoang D. H., Le T. T. D., and Tran T. A. D. Etymology. specific epithet is in reference to the type locality of Bidoup-Nui Ba National Park. Diagnosis. Assigned to the genus Leptolalax on the basis of the following: small size, rounded finger tips, the presence of an elevated inner palmar tubercle not continuous to the thumb, presence of macroglands on body (including supra-axillary, pectoral, femoral and ventrolateral glands), vomerine teeth absent, tubercles on eyelids, anterior tip of snout with vertical white bar (Dubois 1983; Lathrop et al. 1998; Delorme et al. 2006). Leptolalax bidoupensis is distinguished from its congeners by a combination of (1) a dark brownish red ventral surface with white speckling on entire ventral surface including throat, arms and legs, often forming distinct marbling on chest and belly, (2) small size (23.6–24.6 mm in four adult males and 29.2–29.4 mm in two adult females), (3) bicoloured iris (coppery red upper half, fading to pale silver ventrally), (4) a mostly smooth skin texture with no skin ridges, and (5) relatively short tibia (male TIB:SVL 0.44–0.46). The male advertisement call of the new species, consisting of 6–9 single-pulsed notes with a dominant frequency of 1.9–3.8 kHz, is also unique among Leptolala x species for which calls are known. Description of holotype. Head slightly longer than wide; snout bluntly rounded in dorsal view and in profile, projecting slightly beyond margin of the lower jaw; nostril closer to tip of snout than eye; canthus rostralis distinct, gently rounded; lores sloping; vertical pupil; eye diameter smaller than snout length; tympanum distinct, round, diameter smaller than that of the eye; tympanic rim elevated relative to skin of temporal region; vomerine teeth absent; pineal ocellus absent; vocal sac openings oval, located posteriolaterally on floor of mouth; tongue long, moderate width, with slight notch at posterior tip; raised supratympanic ridge running from eye towards axillary gland. Tips of fingers rounded, not swollen; relative finger lengths I 9 %.Published as part of Rowley, Jodi J. L., Le, Duong Thi Thuy, Tran, Dao Thi Anh & Hoang, Huy Duc, 2011, A new species of Leptolalax (Anura: Megophryidae) from southern Vietnam, pp. 15-28 in Zootaxa 2796 on pages 17-26, DOI: 10.5281/zenodo.27701
Comparison of Machine Learning Algorithms for Heartbeat Detection Based on Accelerometric Signals Produced by a Smart Bed
No full textThis work aims to compare the performance of Machine Learning (ML) and Deep Learning (DL) algorithms in detecting users’ heartbeats on a smart bed. Targeting non-intrusive, continuous heart monitoring during sleep time, the smart bed is equipped with a 3D solid-state accelerometer. Acceleration signals are processed through an STM 32-bit microcontroller board and transmitted to a PC for recording. A photoplethysmographic sensor is simultaneously checked for ground truth reference. A dataset has been built, by acquiring measures in a real-world set-up: 10 participants were involved, resulting in 120 min of acceleration traces which were utilized to train and evaluate various Artificial Intelligence (AI) algorithms. The experimental analysis utilizes K-fold cross-validation to ensure robust model testing across different subsets of the dataset. Various ML and DL algorithms are compared, each being trained and tested using the collected data. The Random Forest algorithm exhibited the highest accuracy among all compared models. While it requires longer training time compared to some ML models such as Naïve Bayes, Linear Discrimination Analysis, and K-Nearest Neighbour Classification, it keeps substantially faster than Support Vector Machine and Deep Learning models. The Random Forest model demonstrated robust performance metrics, including recall, precision, F1-scores, macro average, weighted average, and overall accuracy well above 90%. The study highlights the better performance of the Random Forest algorithm for the specific use case, achieving superior accuracy and performance metrics in detecting user heartbeats in comparison to other ML and DL models tested. The drawback of longer training times is not too relevant in the long-term monitoring target scenario, so the Random Forest model stands out as a viable solution for real-time ballistocardiographic heartbeat detection, showcasing potential for healthcare and wellness monitoring applications
Metrological evaluation of contactless sleep position recognition using an accelerometric smart bed and machine learning
No full textPrecise categorization of sleep postures is essential for evaluating overall physical and mental condition. A smart bed was constructed with the microelectromechanical systems (MEMS) accelerometer sensor and an STM 32-bit microcontroller board. This work applies machine learning (ML) methods to acceleration data to accurately categorize four main sleep positions: right side, left side, prone, and supine without any wearable devices. In this work, the efficiency of 9 ML methods is examined. These algorithms include Logistic Regression (LR) with one-vs- rest and multinomial logistic regression types, Linear Discriminant Analysis (LDA), K-Nearest Neighbors Classification (KNN), Classification and Regression Trees (CART), Naive Bayes (NB), Support Vector Machines (SVMs) with one-vs-one and one-vs-rest types, and Random Forest (RF). The best hyperparameters of each model was accomplished, based on GridSearchCV. The K-fold cross-validation with the assessing measurement stability results indicate that the LG-OvR, LDA, and RF models have the best performance, whereas LG-OvR model possesses accuracy rates of almost 99 %. Furthermore, precision, recall and F1-score are calculated with minimum value of 0.95 for all chosen models. The training and test time are also presented for the selected models. This research has important implications for healthcare, sports medicine, and ergonomics, demonstrating the potential of Artificial Intelligence (AI) approaches in improving sleep monitoring methods
Artificial Intelligence Implementation in Internet of Things Embedded System for Real-Time Person Presence in Bed Detection and Sleep Behaviour Monitor
No full textThis paper works on detecting a person in bed for sleep routine and sleep pattern monitoring based on the Micro-Electro-Mechanical Systems (MEMS) accelerometer and Internet of Things (IoT) embedded system board. This work provides sleep information, patient assessment, and elderly care for patients who live alone via tele-distance to doctors or family members. About 216,000 pieces of acceleration data were collected, including three classes: no person in bed, a static laying position, and a moving state for Artificial Intelligence (AI) application. Six well-known Machine-Learning (ML) algorithms were evaluated with precision, recall, F1-score, and accuracy in the workstation before implementing in the STM32-microcontroller for real-time state classification. The four best algorithms were selected to be programmed into the IoT board and applied for real-time testing. The results demonstrate the high accuracy of the ML performance, more than 99%, and the Classification and Regression Tree algorithm is among the best models with a light code size of 1583 bytes. The smart bed information is sent to the IoT dashboard of Node-RED via a Message Queuing Telemetry broker (MQTT)
Leptobrachium leucops Stuart, Rowley, Tran, Le & Hoang, 2011, sp. nov.
No full textLeptobrachium leucops sp. nov. Holotype: NCSM 77465 (field tag BLS 11751; Figures 2, 6), adult male, Vietnam, Langbian Plateau, Lam Dong Province, Lac Duong District, Bidoup-Nui Ba National Park, Hon Giao, 12 ° 11 ' 11.2 "N 108 ° 42 ' 53.5 "E (Figure 7), 1,627 m elev., coll. 6 March 2008 by Bryan L. Stuart, Jodi J. L. Rowley, Tran Thi Anh Dao, Le Thi Thuy Duong, Hoang Duc Huy, Nguyen Le Xuan Bach, and Nguyen Thi Xuan Phuong. Paratypes: Fourteen adult males: AMS R 173163, same data as holotype; FMNH 280396, NCSM 77463 – 64, NCSM 77467, UNS 00121 /AMS R 173159, UNS 00122 /AMS R 173160, UNS 00123 /AMS R 173161, AMS R 173162, AMS R 173165, UNS 00124 /AMS R 173166, AMS R 173168, same data as holotype except coll. 4–16 March 2008; NCSM 77466, AMS R 173164, Vietnam, Langbian Plateau, Khanh Hoa Province, Khanh Vinh District, 12 ° 11 ' 30.5 "N 108 ° 43 '03.5"E, 1,558 m elev., coll. 8 March 2008 by Bryan L. Stuart, Jodi J. L. Rowley, Tran Thi Anh Dao, Le Thi Thuy Duong, Hoang Duc Huy, Nguyen Le Xuan Bach, and Nguyen Thi Xuan Phuong. Two immature females: UNS 00125 /AMS R 173167, same data as holotype except 12 ° 11 ' 33.3 "N 108 ° 42 ' 41.6 "E, 1,900 m elev., coll. 10 March 2008; AMS R 173158, same data as holotype except 12 ° 11 ' 24.3 "N 108 ° 42 '49.0"E, 1,751 m elev., coll. 19 May 2008 by Jodi J. L. Rowley, Tran Thi Anh Dao, Le Thi Thuy Duong, Hoang Duc Huy, Da Du Ha Tien, Vu Hanh Dung, Dinh Binh Phuong, Ly Tri, and Nguyen Thi Xuan Phuong. Etymology. The specific epithet taken from leukos Gr. for white and ops Gr. for eye, in reference to the iris color of the new species. Diagnosis. Assigned to the genus Leptobrachium on the basis of having head width larger than shank length; skin above with a network of ridges; large axillary glands; extremities of digits rounded; and upper part of iris colored differently from lower part (Dubois & Ohler 1998). A small-sized Leptobrachium having males with SVL 38.8–45.2; upper one-third to one-half of iris white; blue scleral arc; dark venter (purplish-gray, dark gray, or black in life, dark gray in preservative) with minute white spots on tubercles; and sexually active males without spines on the upper lip. Description of holotype. Habitus moderately stocky; body tapering to groin. Head broad and depressed; head length and width subequal. Snout rounded in dorsal view, sharply sloping in profile, barely projecting beyond lower jaw in profile; nostril closer to tip of snout than to eye, below canthus, internarial shorter than interorbital distance; canthus rostralis distinct; lores oblique, moderately concave; eye large, slightly projecting from side of head, diameter subequal to snout length, interorbital distance subequal to upper eyelid width; no pineal ocellus; tympanum round, annulus weakly visible, tympanum diameter about 40 % eye diameter and greater than distance between tympanum and eye; tongue heart-shaped, notched posteriorly; large, slit-like vocal sac openings on floor of mouth near lateral margin of tongue; vomerine teeth absent. Forelimb slender. Fingers moderately slender, without webbing. Tip of fingers blunt, those on fingers I and II slightly swollen; relative finger lengths II = IV<I<III; two oval palmar tubercles in contact, inner larger than outer, low callous bumps on ventral surface of fingers; nuptial pad absent. Hindlimb slender and relatively short. Toes moderately slender; webbing on toe I and preaxial side of toe II to level of distal margin of subarticular tubercle, on postaxial side of toe II to base of tip, on preaxial side of toe III to level of proximal subarticular tubercle continuing as a fringe to base of tip, on postaxial side of toe III to midway between proximal subarticular tubercle and tip continuing as a fringe to base of tip, on preaxial and postaxial sides of toe IV to same level as postaxial side of toe III and continuing as a fringe to base of tip, and on toe V to midway between base and tip. Tips of all toes blunt, slightly swollen; relative toe lengths I<II<III = V<IV; distinct, oval, inner metatarsal tubercle, length about 70 % distance between tip of toe I and tubercle; no outer metatarsal tubercle. Skin above smooth with fine network of ridges, with small granules scattered posteriorly, especially near vent; no spines on upper lip; low supratympanic ridge from posterior edge of eye to axilla; ventrally granular, skin smooth on ventral surfaces of limbs; large, round axillary gland on ventrolateral surface slightly posterior to insertion of forelimb with body; oval femoral gland on posteroventral surface of thigh, midway between knee and vent. Color of holotype in life. Dorsum dark gray, with distinct dark brown, Y-shaped marking extending from upper eyelids to lower back, becoming narrower posteriorly, edged with cream, with smaller, irregular dark brown to black markings edged with cream on lower back; upper flank with irregular dark brown and black markings edged with cream or white, lower flank dark gray with minute white spots on tubercles; upper surface of forelimb brown with dark gray bands, each flanked with narrower cream bands; upper surface of hindlimb white or creamywhite with dark grey and black bands; eye black with upper one-third of iris white, scleral arc blue (visible in the posterior corner of the eye and when the palpebrum is retracted); brown bar from edge of upper lip to nostril and from edge of upper lip to lower margin of eye; irregular, black streak under canthus and supratympanic fold, covering tympanum; ventral surface of body and limbs purplish-gray to dark gray, black spotting on chin, minute white spots on tubercles on chin, chest, belly, and banding from upper surfaces of limbs extending to outer margins of lower surfaces of limbs; axillary and femoral glands creamy-white. Color of holotype in preservative. Color in preservative closely resembles color in life, except that cream on dorsal surfaces has faded to white or gray, and purplish-gray on ventral surfaces has faded to dark gray. Measurements of holotype. SVL 41.2; HDL 19.8; HDW 19.5; SNT 7.0; EYE 7.6; IOD 5.3; IND 4.1; SHK 14.6; TGH 18.3; LAL 13.2; HND 9.7; FTL 15.6; IML 2.0; IMW 1.5. Variation. The dorsal pattern is highly variable among paratypes (Figure 2). The background color of the dorsum in life varied from brown to light gray to dark gray. The darker markings on the dorsum are variable in size and shape. One (AMS R 173165) has the dark markings on back arranged as rounded spots (Figure 6). Some (AMS R 173158, UNS 00121 /AMS R 173159, UNS 00123 /AMS R 173161, AMS R 173162 – 64, NCSM 77463, NCSM 77467) have a uniform dark gray dorsum with few or no darker markings (Figure 6). Three paratypes (AMS R 173165, UNS 00124 /AMS R 173166, AMS R 173168) have more distinct limb banding in preservative than the holotype. The ventral coloration is dark gray to black in some paratypes, and some lack black spotting on the chin. The white coloration in the eye extends to about one half of the iris in some paratypes. Measurements are summarized in Table 1. Advertisement call. The following description is based on the calls of five individuals recorded at 13.0– 18.4 °C (Table 4). The call contains 1–5 (but usually 3–4), highly-pulsed notes and lasted 0.10– 1.70 s, repeated at a variable interval (6.5– 124.1 s). Within a call, notes were relatively evenly spaced and each note contained 8–15 pulses repeated at a near constant rate across the call. Relative amplitude varied symmetrically within each note, gradually rising over the first half of each note and then declining gradually across the second half of the note (Figure 5). The dominant frequency of calls varied from 1.0– 1.6 kHz, and spanned approximately 1 kHz (Figure 5). A low frequency band was present at around 0.2 kHz, and a weak harmonic was present at approximately 2.6 kHz. There was slight frequency modulation within each note, with the first few pulses and occasionally the last few pulses approximately 0.1 kHz lower in frequency. Temporal and spectral properties of the call did not obviously vary with temperature in the five specimens recorded (Table 4). To the human ear, the call sounds like a rapid barking “wah-wah-wah-wah”. Distribution and natural history. Leptobrachium leucops is known only from 1,558–1,900 m elevation on the Langbian Plateau in Lam Dong and Khanh Hoa Provinces, Vietnam. The species occurs in wet evergreen and cloud forest, where males call from shallow burrows under leaf litter. Leptobrachium leucops occurs in syntopy with L. pullum at some sites; in March, males of both species were heard calling within a few meters of each other. Comparisons. Only two other named species of Leptobrachium that occur in Vietnam, Laos, or Cambodia (Table 2) have the upper part of the iris white: L. banae Lathrop, Murphy, Orlov & Ho, 1998 and L. xanthospilum Lathrop, Murphy, Orlov & Ho, 1998. Both species are restricted to the Kon Tum Plateau of Vietnam (Lathrop et al. 1998), the nearest uplands to the Langbian Plateau (Figure 7). However, both are considerably larger than the new species, with SVL of males in L. xanthospilum 62.8–73.4 and L. banae 57.2 –70.0 (38.8–45.2 in L. leucops). Leptobrachium xanthospilum further differs by having distinct, large, yellow, glandular spots on the flank (absent in L. leucops). Leptobrachium banae further differs by having red bands on the limbs (absent in L. leucops) and a white scleral arc (blue in L. leucops). Leptobrachium chapaense (Bourret, 1937) has been reported to have a black iris (upper and lower parts not differing) or the upper part of the iris white or blue (Dubois & Ohler 1998; Lathrop et al. 1998). Molecular evidence suggests that L. chapaense is actually a complex of species across its range (Rao & Wilkinson 2008); at its type locality at Sa Pa in northwestern Vietnam, the eyes are black (Bourret, 1937; Orlov, 2005). Leptobrachium “ chapaense ” at Tam Dao, northern Vietnam, has a white upper iris, but is larger than the new species, with SVL of males 53.5–65.5 (38.8–45.2 in L. leucops), and has orange blotches on the sacral region, flank, and dorsal surfaces of limbs (absent in L. leucops; Lathrop et al. 1998). The eye color of L. ngoclinhense (Orlov, 2005), unreported in its original description (Orlov, 2005), is dark brown to black (no light-colored upper iris) with a white scleral arc (J. J. L. Rowley, unpublished). Parameter IndividualPublished as part of Stuart, Bryan L., Rowley, Jodi J. L., Tran, Dao Thi Anh, Le, Duong Thi Thuy & Hoang, Huy Duc, 2011, The Leptobrachium (Anura: Megophryidae) of the Langbian Plateau, southern Vietnam, with description of a new species, pp. 25-40 in Zootaxa 2804 on pages 32-36, DOI: 10.5281/zenodo.20815
Deep Learning Approach to LPI Radar Recognition
No full textIn this study, an advanced automatic low probability of intercept (LPI) radar recognition technique (LWRT) that includes both LPI radar signal classification and parameter extraction is proposed. It is shown with Monte Carlo simulation that, even without the unrealistic assumptions used in the previous studies, the proposed LWRT achieves classification performance similar to that of the state-of-the-art LWRT for pulse wave (PW) LPI radar waveforms. And by the combination of the 'single shot multi-box detector' (SSD) or 'you only look once version 3' (YOLOv3) and a supplementary classifier, the proposed LWRT achieves an extraordinary classification performance for continuous (CW) LPI radar waveforms for all the twelve modulation schemes considered in the literature (i.e., BPSK, Costas, LFM, Frank, P1, P2, P3, P4, T1, T2, T3, and T4). Moreover, the proposed LWRT summarizes the existing and proposed new parameter extraction functions, which can help to design the countermeasure in electronic warfare. © 2019 IEEE
- …
