Image dehazing is an ill-posed problem that has been extensively studied in the recent years. The objective performance evaluation of the dehazing methods is one of the major obstacles due to the lacking of a reference dataset. While the synthetic datasets have shown important limitations, the few realistic datasets introduced recently assume homogeneous haze over the entire scene. Since in many real cases haze is not uniformly distributed we introduce NH-HAZE, a non-homogeneous realistic dataset with pairs of real hazy and corresponding haze-free images. This is the first non-homogeneous image dehazing dataset and contains 55 outdoor scenes. The non-homogeneous haze has been introduced in the scene using a professional haze generator that imitates the real conditions of hazy scenes. Additionally, this work presents an objective assessment of several state-of-the-art single image dehazing methods that were evaluated using NH-HAZE dataset.

Real-world application scenarios like modern Internet services consist of diversity of AI and non-AI modules with very long and complex execution paths. Using component or micro AI benchmarks alone can lead to error-prone conclusions. This paper proposes a scenario-distilling AI benchmarking methodology. Instead of using real-world applications, we propose the permutations of essential AI and non-AI tasks as a scenario-distilling benchmark. We consider scenario-distilling benchmarks, component and micro benchmarks as three indispensable parts of a benchmark suite. Together with seventeen industry partners, we identify nine important real-world application scenarios. We design and implement a highly extensible, configurable, and flexible benchmark framework. On the basis of the framework, we propose the guideline for building scenario-distilling benchmarks, and present two Internet service AI ones. The preliminary evaluation shows the advantage of scenario-distilling AI benchmarking against using component or micro AI benchmarks alone. The specifications, source code, testbed, and results are publicly available from the web site url{this http URL}.

We present the Perceptimatic English Benchmark, an open experimental benchmark for evaluating quantitative models of speech perception in English. The benchmark consists of ABX stimuli along with the responses of 91 American English-speaking listeners. The stimuli test discrimination of a large number of English and French phonemic contrasts. They are extracted directly from corpora of read speech, making them appropriate for evaluating statistical acoustic models (such as those used in automatic speech recognition) trained on typical speech data sets. We show that phone discrimination is correlated with several types of models, and give recommendations for researchers seeking easily calculated norms of acoustic distance on experimental stimuli. We show that DeepSpeech, a standard English speech recognizer, is more specialized on English phoneme discrimination than English listeners, and is poorly correlated with their behaviour, even though it yields a low error on the decision task given to humans.

Digital watermarking is a remarkable issue in the field of information security to avoid the misuse of images in multimedia networks. Although access to unauthorized persons can be prevented through cryptography, it cannot be simultaneously used for copyright protection or content authentication with the preservation of image integrity. Hence, this paper presents an optimized multipurpose blind watermarking in Shearlet domain with the help of smart algorithms including MLP and NSGA-II. In this method, four copies of the robust copyright logo are embedded in the approximate coefficients of Shearlet by using an effective quantization technique. Furthermore, an embedded random sequence as a semi-fragile authentication mark is effectively extracted from details by the neural network. Due to performing an effective optimization algorithm for selecting optimum embedding thresholds, and also distinguishing the texture of blocks, the imperceptibility and robustness have been preserved. The experimental results reveal the superiority of the scheme with regard to the quality of watermarked images and robustness against hybrid attacks over other state-of-the-art schemes. The average PSNR and SSIM of the dual watermarked images are 38 dB and 0.95, respectively; Besides, it can effectively extract the copyright logo and locates forgery regions under severe attacks with satisfactory accuracy.

We study the specification and verification of hyperproperties for probabilistic systems represented as Markov decision processes (MDPs). Hyperproperties are system properties that describe the correctness of a system as a relation between multiple executions. Hyperproperties generalize trace properties and include information-flow security requirements, like noninterference, as well as requirements like symmetry, partial observation, robustness, and fault tolerance. We introduce the temporal logic PHL, which extends classic probabilistic logics with quantification over schedulers and traces. PHL can express a wide range of hyperproperties for probabilistic systems, including both classical applications, such as differential privacy, and novel applications in areas such as robotics and planning. While the model checking problem for PHL is in general undecidable, we provide methods both for proving and for refuting a class of probabilistic hyperproperties for MDPs.

Recent works imply that the channel pruning can be regarded as searching optimal sub-structure from unpruned networks.

However, existing works based on this observation require training and evaluating a large number of structures, which limits their application.

In this paper, we propose a novel differentiable method for channel pruning, named Differentiable Markov Channel Pruning (DMCP), to efficiently search the optimal sub-structure.

Our method is differentiable and can be directly optimized by gradient descent with respect to standard task loss and budget regularization (e.g. FLOPs constraint).

In DMCP, we model the channel pruning as a Markov process, in which each state represents for retaining the corresponding channel during pruning, and transitions between states denote the pruning process.

In the end, our method is able to implicitly select the proper number of channels in each layer by the Markov process with optimized transitions. To validate the effectiveness of our method, we perform extensive experiments on Imagenet with ResNet and MobilenetV2.

Results show our method can achieve consistent improvement than state-of-the-art pruning methods in various FLOPs settings. The code is available at https://github.com/zx55/dmcp

A big part of the video content we consume from video providers consists of genres featuring low-light aesthetics. Low light sequences have special characteristics, such as spatio-temporal varying acquisition noise and light flickering, that make the encoding process challenging. To deal with the spatio-temporal incoherent noise, higher bitrates are used to achieve high objective quality. Additionally, the quality assessment metrics and methods have not been designed, trained or tested for this type of content. This has inspired us to trigger research in that area and propose a Grand Challenge on encoding low-light video sequences. In this paper, we present an overview of the proposed challenge, and test state-of-the-art methods that will be part of the benchmark methods at the stage of the participants' deliverable assessment. From this exploration, our results show that VVC already achieves a high performance compared to simply denoising the video source prior to encoding. Moreover, the quality of the video streams can be further improved by employing a post-processing image enhancement method.

Over the years, performance evaluation has become essential in computer vision, enabling tangible progress in many sub-fields. While talking-head video generation has become an emerging research topic, existing evaluations on this topic present many limitations. For example, most approaches use human subjects (e.g., via Amazon MTurk) to evaluate their research claims directly. This subjective evaluation is cumbersome, unreproducible, and may impend the evolution of new research. In this work, we present a carefully-designed benchmark for evaluating talking-head video generation with standardized dataset pre-processing strategies. As for evaluation, we either propose new metrics or select the most appropriate ones to evaluate results in what we consider as desired properties for a good talking-head video, namely, identity preserving, lip synchronization, high video quality, and natural-spontaneous motion. By conducting a thoughtful analysis across several state-of-the-art talking-head generation approaches, we aim to uncover the merits and drawbacks of current methods and point out promising directions for future work. All the evaluation code is available at: https://github.com/lelechen63/talking-head-generation-survey.

This paper investigates a class of optimal control problems associated with Markov processes with local state information. The decision-maker has only local access to a subset of a state vector information as often encountered in decentralized control problems in multi-agent systems. Under this information structure, part of the state vector cannot be observed. We leverage ab initio principles and find a new form of Bellman equations to characterize the optimal policies of the control problem under local information structures. The dynamic programming solutions feature a mixture of dynamics associated unobservable state components and the local state-feedback policy based on the observable local information. We further characterize the optimal local-state feedback policy using linear programming methods. To reduce the computational complexity of the optimal policy, we propose an approximate algorithm based on virtual beliefs to find a sub-optimal policy. We show the performance bounds on the sub-optimal solution and corroborate the results with numerical case studies.

When it comes to the music scene in the midst of the coronavirus pandemic, the math is pretty simple: No shows equals no revenue.…

Put pictures of 1975-era Aerosmith and 1995-era Guided By Voices next to each other and you probably wouldn’t think the bands have anything in common.…

Every serious pinball player remembers their first machine.…

A parking assist system includes: an actuator that drives a back door of a vehicle; an opening degree control unit that controls an opening degree of the back door by controlling the actuator; a storage device that stores an allowable opening degree of the back door at a park position of the vehicle in association with the park position; and a position information acquisition unit that acquires position information of the vehicle. When a position of the vehicle corresponds to the park position stored in the storage device, the opening degree control unit limits the opening degree of the back door on the basis of the allowable opening degree of the back door, stored in the storage device in association with the park position.

Methods and apparatus to generate and use content-aware watermarks are disclosed herein. In a disclosed example method, media composition data is received and at least one word present in an audio track of the media composition data is selected. The word is then located in a watermark.

A message screen display comprises a static non-scrollable display area for display of at least part of a first message, the first message having an associated first message time. The message screen display further comprises a scrollable display area for display of at least part of a second message, the second message having an associated second message time. The message screen display further comprises a feature applied to at least part of the first message that varies based on time as referenced to the associated first message time.

PCT No. PCT/CH96/00292 Sec. 371 Date May 22, 1998 Sec. 102(e) Date May 22, 1998 PCT Filed Aug. 23, 1996 PCT Pub. No. WO97/08592 PCT Pub. Date Mar. 6, 1997A method by which an elastomer is applied to a model (1) with relief marks to be reproduced. The purpose of the elastomer once set is to form a mold membrane (14) and incorporate within its own material the shape of the marks on the model. The mold membrane is then separated from the model and filled with a setting material, and the mold membrane is applied to a support in order to transfer the marks after preliminary pasting.

An intracorporeal marker, for marking a site within living tissue of a host, includes an outer body portion of biodegradable material. An inner body portion is located in the outer body portion. The inner body portion includes biological material that becomes calcified in the living tissue of the host over time. An agent interacts with the biological material to promote calcification of the biological material of the inner body portion in the living tissue of the host.

The methods described herein enable the evaluation of compounds on subjects to assess their therapeutic efficacy or toxic effects. The target of analysis is the underlying biochemical process or processes (i.e., metabolic process) thought to be involved in disease pathogenesis. Molecular flux rates within the one or more biochemical processes serve as biomarkers and are quantitated and compared with the molecular flux rates (i.e., biomarker) from control subjects (i.e., subjects not exposed to the compounds). Any change in the biomarker in the subject relative to the biomarker in the control subject provides information to evaluate therapeutic efficacy of an administered drug or a toxic effect and to develop the compound further if desired. In one aspect of the invention, stable isotope-labeled substrate molecules are administered to a subject and the label is incorporated into targeted molecules in a manner that reveals molecular flux rates through metabolic pathways of interest.

A color-changing marking system, including a color-changing marking medium and a substrate. The color-changing marking medium has a nominal color and includes a color developer activator and/or color former leuco dye. The substrate includes a treated portion on which a color former leuco dye and/or color developer activator has been applied. Application of the color-changing marking medium to an untreated portion of the substrate causes the color-changing marking medium to create marks of the nominal color. Application of the color-changing marking medium to the treated portion of the substrate causes the color-changing marking medium to create marks of a changed color in response to a chemical or mechanical reaction between the leuco dye and the activator. The marking medium may a crayon, colored pencil, marker ink or paint.

A method of color laser marking an article having a polymeric foil with at least one colorless layer containing an infrared absorber, a polymeric binder and a color forming compound; including the steps of:—laser marking the colorless layer with an infrared laser using a first laser operation mode to generate a blue or cyan color; and—laser marking the same colorless layer with an infrared laser using a second laser operation mode to generate a black color, wherein the first laser operation mode applies less energy to the colorless layer than the second laser operation mode. Also disclosed is an article, such as a security document, including a polymeric foil and a colorless layer containing laser marked graphical data having a blue or cyan color and laser marked information having a black color.

A security document precursor including, in order: a) at least one transparent biaxially stretched polyester foil; b) one colorless color forming layer containing at least an infrared absorber, a colorless dye-precursor and a polymeric binder; and c) a polymeric support; wherein the polymeric binder is copolymer including at least 90 wt % of a chlorinated ethylene and 1 wt % to 10 wt % of vinyl acetate both based on the total weight of the binder. Methods for making the security document precursor are also disclosed.

A method for preparing a color laser marked article comprising the steps of: a) infrared laser marking a security element including a polymeric support and a color forming layer comprising a color forming compound, an infrared dye and a polymeric binder comprising vinyl acetate and at least 85 wt % of vinyl chloride based on the total weight of the binder; and b) exposing the laser marked security element with light having a wavelength higher than 440 nm. The light exposure of step b) hinders the falsification of an issued security document without significant increase of background density.

A security document precursor including, in order, at least: a) a transparent biaxially stretched polyethylene terephthalate foil; b) a colorless color forming layer containing at least an infrared absorber, a color forming component and a polymeric binder; and c) a polymeric support; wherein the colorless color forming layer contains at least one component forming a compound having a melting temperature of less than 20° C. upon laser marking the colorless color forming layer with an infrared laser. Methods for securing a security document using the security document precursor are also disclosed.

Various embodiments of stents with radiopaque markers arranged in patterns are described herein.

An implant for marking an area within a living body includes a matrix material and a marking material. The implant is formable to fit the shape and size of a cavity in the human body. The implant is configured to support tissue surrounding the cavity and to allow in-growth of fibrous tissue into and replace at least a portion of the matrix material.

A starting device of a spark-ignition multi-cylinder engine is provided. The device includes a multi-cylinder engine body having cylinders, fuel injection valves, ignition plugs, an intake valve drive mechanism for opening and closing intake valves, a hydraulic variable valve phase mechanism for changing a close timing of each intake valve, an engine-driven hydraulic pressure supply source for supplying a hydraulic pressure, and a start controller for controlling the fuel injection valves, the ignition plugs, and the variable valve phase mechanism. When the supplied hydraulic pressure is below a predetermined pressure, the variable valve phase mechanism locks the close timing. When an engine temperature in an engine-start is high, the start controller retards a fuel injection timing of the cylinder on intake stroke at an engine stopped timing and retards an ignition timing thereof. The start controller does not retard the close timing of the intake valve until the engine-start completes.

The present invention provides a family of dark quenchers, termed Black Hole Quenchers (“BHQs”), that are efficient quenchers of excited state energy but which are themselves substantially non-fluorescent. Also provided are methods of using the BHQs, probes incorporating the BHQs and methods of using the probes.

A road marker and related light based warning device are described. The road marker or device includes a thermal sensor that triggers the illumination of at least one light-emitting diode at a predetermined temperature. The temperature may be associated with ice formation. The light-emitting diode(s) may flash to alert motorists to hazardous road conditions. The road marker or device utilize simple components to increase reliability, particularly when the device is subjected to high heat such as when the device is mounted into hot tar seal. The device further includes a switching element that prevents rapid on/off cycling.

A mobile thermoplastic applicator comprises a kettle, a main gas burner, a heat exchanger plenum, and a screed die box. The kettle has walls which define an interior space for melting thermoplastic. The main gas burner is situated beneath the kettle, and is configured to heat the kettle. The heat exchanger plenum is situated inside the thermoplastic kettle and is configured to subdivide the interior space and circulate heated air from the main gas burner. The screed die box is configured to receive molten thermoplastic from the kettle and deposit that molten thermoplastic on a surface.

The invention relates to a marking apparatus (10) for marking an object comprising a marking head (20) having a plurality of marking devices (40, 40a, 40b) for applying a marking on the object and a driving mechanism for providing a relative movement of the object relative to the marking head in an advance direction (16) during a marking operation. The marking head comprises in addition to the plurality of marking devices a plurality of sensor devices and the sensor devices (50) are arranged down-stream of the marking devices in the advance direction, so that the marking applied by the marking devices is detectable by the sensor devices, when the object is moved relative to the marking head in the advance direction. The invention also relates to a method for marking an object.

A positive response notification to provide information regarding locate and/or marking operations for underground facilities may include time-stamp information to provide proof of a time at which the locate and/or marking operation was completed by a locate technician, and/or place-stamp information to provide proof of a presence of the locate technician at or near a work site. An electronic manifest image and/or a virtual white line image similarly may be included in a positive response notification. In one example, such images may be bundled together based on respective descriptor files (or descriptor metadata) that associates the corresponding images with a locate request ticket for the operation. In another example, a positive response notification may include environmental information regarding one or more environmental conditions present at or near the work site during the locate and/or marking operation.

A method is provided for operating a parking brake module that is at least partially integrated into a compressed air generation system in the event of defects, having a control unit, solenoid valves, and a relay valve for aerating and deaerating at least one spring-loaded brake cylinder. A pressure in the parking brake module which is elevated compared to a normal pressure is determined. A constant compressed air delivery is stopped. A reduced switch-off pressure of the compressed air generation system is set. The pressure level in the parking brake module is lowered to the reduced switch-off pressure through repeated activation of the relay valve.

A method for hydraulically boosting a vehicle electric parking brake having a hydraulic service brake and an electric parking brake. Application forces electically generated by the parking brake function is superimposed on the boosting brake force generated by a hydraulic boosting brake pressure provided by the service brake to the brake actuator. When the parking brake is actuated to generate a predetermined application force, the force generated by the brake actuator is detected as the actual value, if an actual value of the measured value is smaller than a first target value a boosting brake pressure value is applied to the brake actuator, and by means of the parking brake function an adjustment function for the tension of the brake actuator to which the hydraulic boosting brake pressure is applied is carried out to achieve the predetermined application force.

Disclosed is a control method of an electronic parking brake system, which variably controls the duty of voltage applied to a motor upon release of the electronic parking brake system. The control method includes controlling voltage applied to a motor to a first duty ratio upon release of the electronic parking brake system, controlling the voltage applied to the motor to a second duty ratio greater than the first duty ratio if locking of the motor occurs and the motor is not operated, and controlling the voltage applied to the motor to the first duty ratio if locking of the motor is released and the motor begins to operate, after the control of voltage to the second duty ratio.

Parking arrangement and method for making a parking arrangement equipped with an automatic vehicle detection system ready for operation, which parking arrangement comprises a central computer with a database and at least a number of parking places identifiable by a location code, which parking places are each provided with at least one wirelessly operating parking sensor module, which is connected with the central computer via a UHF radio link and which is provided with an identification code, which parking sensor module comprises at least one vehicle sensor which in operation provides measuring values representative of the presence or absence of a vehicle in the respective parking place, wherein use is made of parking sensor modules which have an RFID identification circuit in which the identification code is stored, which identification code is wirelessly readable.

A method and device for activating and deactivating a valet parking function for a motor vehicle is provided. An activated valet parking function prevents access to one or more lockable areas in the motor vehicle. The method including the steps of checking whether a freely operable control element has been actuated. If the actuation of a control element has been detected, the valet parking function is activated to prevent exclusively the access to the lockable area in the motor vehicle, particularly to a trunk.

A parking assistance apparatus has a region detector capable of detecting a plurality of possible parking regions where a vehicle can park in at least one side to the left or right of the vehicle, a photographic image of the scenery surrounding the vehicle including the possible parking regions is displayed on a monitor device, the possible parking regions detected by the region detector are superimposed on the photographic image as graphic images showing parking target candidates in all of the corresponding positions, and one of the parking target candidates can be selected as a parking target on the basis of a command input from a driving device of the vehicle.

The invention describes biomarkers which can be used to predict the likelihood that an individual will develop Diabetes. The biomarkers can also be used to screen large groups in order to identify individuals at risk of developing Diabetes.

This invention relates, e.g., to a method for determining if a subject has myocardial ischemia, comprising (a) providing a blood sample obtained from a subject suspected of having myocardial ischemia; (b) determining in the sample the amount of one or more of the following proteins: (i) Lumican and/or (ii) Extracellular matrix protein 1 and/or (iii) Carboxypeptidase N; and (c) comparing the amount(s) of the protein(s) to a baseline value that is indicative of the amount of the protein in a subject that does not have myocardial ischemia, wherein a statistically significantly increased amount of the protein(s) compared to the baseline value is indicative of myocardial ischemia. Other proteins indicative of myocardial ischemia are also described, as are methods for treating a subject based on a diagnostic procedure of the invention, and kits for carrying out a method of the invention.

The present disclosure is related to illustrative embodiments of an apparatus and method for maintaining carbonation of sparkling beverages, such as sparkling wines, cocktails, and non-alcoholic beverages such as sodas and coffee. A transparent, shatter-resistant, height-adjustable carbonation enclosure is used to contain an opened bottle of a sparkling beverage. An air-tight carbonation cap with a built-in one-way valve is used to re-carbonate the beverage inside the bottle using a pressurized gas source. The base of the carbonation enclosure includes a platform that may be used to adjust the effective height of the enclosure to fit the height of the bottle and provide an air-tight interface between an opening of the bottle and the carbonation cap. The carbonation enclosure further includes a number of openings to open air to prevent pressurization of the carbonation enclosure. Another illustrative embodiment includes a safety enclosure for multiple bottles pressurized directly by conical nozzles.

A print drum subassembly suitable for ink jet printing applications includes a composite print drum including an outer shell having a wall thickness in a range of about 1.5 mm to about 15 mm and a thermal conductivity greater than about 200 W/m-K disposed around a hollow core having wall thickness in a range of about 4 mm to about 30 mm and a thermal conductivity less than about 10 W/m-K. A radiant heater is arranged within the hollow core and is configured to heat the outer shell without substantially heating the hollow core.

An inner peripheral surface of a tube-shaped inner casing has a regular hexagonal shape when viewed from the axial direction, a plurality of pins are arranged in the inner casing so as to take a posture parallel to the axial direction of the inner casing, and in this arrangement, a plurality of rows of the pins arranged in a predetermined direction are arranged in a direction perpendicular to the row direction of the pins, the adjacent pins of the adjacent rows are offset from each other in the row direction of the pins, and each pin contacts the pins adjacent thereto.

A marking head of a marking device comprises: at least one marking pin; a heater that heats the marking pin; a support base that supports a first marking tape and a second marking tape in such a manner that the first marking tape and the second marking tape line up in a direction orthogonal to the longitudinal direction of the marketing pin; a driving device that is capable of moving the support base to a first position at which the first marking tape faces the marking pin, and a second position at which the second marking tape faces the marking pin; and extrusion devices that extrude the marking pin out toward the marking tape facing the marking pin in order to press the marking tape facing the marking pin against a tire and print on the tire.

A board game system comprises one or more game objects, a processing device, a memory device and one or more cameras. Each of the game objects comprise a unique visual marker positioned on a top surface of the game object, wherein the unique visual marker comprises a series of concentric rings that represent data that uniquely identifies the game object. As a result, during the course of game play, the location and identification of the game objects are able to be determined by the processing device by analyzing images captured by the one or more cameras of the visual markers of the game objects on the game board. The processing device is able to compare the data of the visual markers to a table stored in the memory device that associates the data with a specific game object.

The present invention relates to a security device for improving the marketing impact of an article of commerce. The device generally includes a housing having a base and a cover with the cover having at least one opening to allow for the insertion of an attachment mechanism. The housing element of the present invention in one embodiment is in a shape representing a particular theme such as a particular trademark, trade dress, special events, commemorative events, historical events and combinations thereof. The housing may contain an RFID device and/or an electronic article security device. The present invention also provides a method of attaching a security device described herein to an article of commerce.

A technique of reducing occurrence of multiple discharge in a spark plug. The spark plug has a main ground electrode and three auxiliary ground electrodes. The position at which first auxiliary ground electrode is joined to a metallic shell is located opposite the position at which main ground electrode is joined to the metallic shell, with respect to a center electrode. The positions at which second and third auxiliary ground electrodes are joined to the metallic shell are located opposite to each other with respect to the center electrode. When the width of first auxiliary ground electrode is represented by W, the shortest distance between second auxiliary ground electrode and third auxiliary ground electrode is represented by T, and a distance which is a component of the shortest distance T in a direction orthogonal to first auxiliary ground electrode is represented by Tp, a relation W≧Tp is satisfied.

The spark plug has a configuration satisfying the relationships of B≧0.7A and 0.3 mm≦A≦0.6 mm, where B is an axial thickness along the central axis line Q of the weld portion formed between the base material electrode and the noble-metal chip, and A is an axial distance along the central axis line Q between the intersection points P3 and X. The intersection point P3 is a point at which a phantom axis line radially distant from the central axis line Q by D/2 (D being a diameter of the noble-metal chip) intersects with the boundary line between the weld portion and the noble-metal chip. The intersection point X is a point at which an extension of the contour line of the base material electrode in the vicinity of the weld portion intersects with a boundary line between the weld portion and the base material electrode.

A method of making a spark plug electrode includes several steps. One step includes providing an inner core of a ruthenium (Ru) based alloy or an iridium (Ir) based alloy. Another step includes providing an outer skin over a portion or more of the inner core in order to produce a core and skin assembly. The outer skin can be made of platinum (Pt), gold (Au), silver (Ag), nickel (Ni), or an alloy of one of these. Yet another step includes increasing the temperature of the core and skin assembly. And another step includes hot forming the core and skin assembly at the increased temperature.

An industrial spark plug (20) includes a central electrode (24) with a central base (30) formed of a nickel-based material and a central firing tip (32) formed of an iridium-based material. The central firing tip (32) has a tip thickness (tct) of 0.02 to 0.03 inches, a tip diameter (dct) of 0.1184 to 0.1776 inches, and an aspect ratio of 4.736 to 7.104. The central firing tip (32) is electron beam welded to the central base (30) to provide a robust joint therebetween. The central electron beam weld (36) includes a mixture of re-crystallized iridium-based material and re-crystallized nickel-based material extending continuously along and over the entire welding interface. The spark plug (20) also includes a ground electrode (26) with a ground firing tip (38) electron beam welded to a ground base (42).

A rail seat includes a concrete rail tie; a pair of rail clip support shoulders cast in place in said rail tie each said shoulder having a rail face between its external sides; a pair of shoulder insulators each shaped to lie against the rail face of the rail shoulder; a pair of rail clips each having a base section adapted to seat within the rail clip support shoulder and a toe section adapted to seat on the rail base. The clip and toe insulator assembly is loaded on the shoulder in a vertical direction so that parts of the toe insulator seat within pockets in the shoulder that enable the clip and insulator assembly to move laterally towards the rail but not away from it and the clip base is aligned with the shoulder slot ready to enter it when the clip is moved laterally to the parked position which applies a small downward force to the clip toes and at this stage lie outside the limits of the rail base and finally the clip can be moved laterally into the installed position where the toe insulators sit on top of the rail base and securely hold it down.