28 Mar 2018

Castagnet Claims Wimbledon Club Squash Squared Open Title

 

France’s World No.28 Mathieu Castagnet claimed his first title since the 2016 Canary Wharf Classic at the Wimbledon Club Squash Squared Open as the fifth seed made it two final defeats in a row for England’s World No.34 Tom Richards.
Castagnet – who defeat Ben Coleman, Youssef Soliman and James Willstrop en route to the final – took control of the opening game to take an important lead.
The second game was a much closer affair with Richards – who was beaten in the final last year by Alan Clyne – executing his shots better, however, the Frenchman was able to get the better of his English opponent in a close ending and then recovered from an early 5-0 deficit to take the third and complete a straight-games victory by an 11-6, 11-9, 11-6 scoreline.

The win marks Castagnet’s first PSA World Tour title since the Canary Wharf Classic back in March 2016.
“The last time I won one of the biggest tournaments in my career was in London at Canary Wharf,” said the 31-year-old Frenchman.
“And two years later I have won this one – I should be born here!
“I want to thank my wife, every day she helps me make my comeback on the PSA World Tour and so I really want to thank her and Arnaud Hays and Renan Lavigne.
“I think the difference was the second game, when I got it to go 2-0 up it gave me so much confidence because you just need one game and he needs three and it’s hard to come back.
“He started the third game well and I saw him struggling a bit when I was using my cross, so I kept pushing him.”
Result – Final
[5] Mathieu Castagnet (FRA) bt [7] Tom Richards (ENG) 3-0: 11-6, 11-9, 11-6

Sobhy lands Texas Open crown


Amanda Sobhy continued her strong return from injury with victory in the final of the Novum Energy Texas Open Women’s Squash Championship.
The World No.18, who suffered an Achilles injury a year ago which saw her ranking fall out of the top ten, defeated fellow American Reeham Sedky 11-8 10-12 11-8 11-9 in the first all-American final in Houston.
Qualifier Sedky had enjoyed an impressive tournament, knocking out top seed Hania El Hammamy in round one and then backed this up with wins over number seven seed Danielle Letourneau and number four seed Mayar Hany.
Sobhy, meanwhile, had only dropped one game en route to the final as she dispatched Finland’s Emilia Soini, Egypt’s Rowan Elaraby and Nada Abbas.
Sobhy took the opening game before Sedky battled back to level the scores in the second on the tie-break. However, she could not hold off her compatriot who completed a 3-1 victory in 47 minutes.
Results – final
[3] Amanda Sobhy (USA) bt [Q] Reeham Sedky (USA) 3-1: 11-8 10-12 11-8 11-9 (47m)
Amanda Sobhy with the 2018 Texas Open trophy

PSA Unveil New Tour Structure for Professional Squash and Announce Initiative with World Squash Federation


 
The Professional Squash Association (PSA) today unveiled its new vision for professional squash which will see a new-look professional squash tour structure come into effect for the 2018/19 season onwards, while a WSF & PSA Satellite Tour has also been announced following the recent partnership agreement between the PSA and World Squash Federation (WSF).
 
The changes, which have been made in order to streamline the sport and increase earning potential for professional squash players, will see the current PSA World Tour rebranded into two individual circuits - comprising of the PSA World Tour and PSA Challenger Tour - while qualification rounds will be scrapped and a series of new tournament tiers will be introduced across both circuits. 
 
Under the new look, the PSA World Tour will continue to be home to all current top-tier events offering total prize money from $50,000 - $1,000,000 and above including the PSA World Championships and PSA World Tour Finals - while new tournament tiers in the form of World Tour Platinum, featuring 48-player draws, and World Tour Gold, World Tour Silver and World Tour Bronze, all of which are set to feature 24-player draws, will be introduced.
 
The 2018/19 PSA World Championships, which will be held in Chicago with a record $1 million prize fund up for grabs, will be the first World Championships to be staged under the new tour structure. Both draws will comprise of 56 PSA entrants and one wildcard, with the other seven places taken up by winners from selected tournaments on the PSA Challenger Tour in a 'Road to Chicago' competition.
 
The PSA Challenger Tour will provide a platform for the world's most exciting up-and-coming players to cut their teeth on the professional circuit against some of squash's most experienced professionals, with tournaments offering prize money between $5,500 - $30,000. Tournament tiers consist of Challenger Tour 30, Challenger Tour 20, Challenger Tour 10 and Challenger Tour 5 tournaments, while a round robin system will be trialled at selected Challenger Tour 5 tournaments throughout the 2018/19 season.
 
Following on from last year’s partnership agreement between the PSA and the World Squash Federation (WSF), the WSF & PSA Satellite Tour will be formed, which will integrate sanctioned World, Regional, National and Junior tournaments into the PSA World Rankings to create a pathway to the professional circuit for aspiring, up-and-coming players, strengthening the link between the global governing body for squash and the professional tour.
 
Comprising the WSF National Championships, Satellite Tournaments, World & Regional Junior Championships, National Junior Opens and National Junior Championships, the WSF & PSA Satellite Tour will run throughout each calendar year.
 
Players will earn PSA World Ranking points based on their results at tournaments on the PSA World Tour, PSA Challenger Tour and the WSF & PSA Satellite Tour.
 
"We are committed to promoting a thriving and sustainable professional squash circuit and feel that simplifying the tour structure and easing the transition from junior squash to professional squash will help us to achieve those aims," said PSA Chief Executive Alex Gough.
 
"Since signing a Memorandum of Understanding with the WSF in August 2017, we have been hugely encouraged by the commitment from both sides to forge a strong working relationship and believe that partnering with the WSF in this manner will be a significant boost to the growth of the sport in the coming years."
 
PSA Tour Director Hannah Ridgard-Mason said: "This is an exciting time for the tour and we believe that these changes will enable us to further promote the game at the highest level while creating visibility for the PSA Challenger Tour, which will be the backbone of the new tour structure.
 
"We would like to thank our players and tournament promoters for their cooperation during this transition and look forward to working with them closely to ensure that the new tour structure will be a success."
 
WSF President Jacques Fontaine said: "It was an important intention for me to bring WSF and PSA closer together when I became WSF President, as a unified sport will progress even more strongly. Our MoU has achieved this aim. Now joining together to allow players competing in our member nation's national and junior championships to have the opportunity of ranking points is an example showing the benefits of this unity.    
 
"Our priorities as a sport are to offer leadership in education, accessibility and sustainability, and competition too – and our strong bond with PSA addresses the last pillar especially. I am sure that the WSF & PSA Satellite Tour will go from strength to strength."


About PSA
The Professional Squash Association (PSA) is the global governing body responsible for the administration of both men’s and women’s professional squash around the world.
With over 850 registered players and more than 200 events taking place around the globe every year, the PSA World Tour – managed and run by the PSA – showcases the game of squash at its highest level in some of the world’s most unique and stunning locations.
Encapsulating all that is good about top flight professional squash and combining athleticism, skill, fitness, tenacity, power, drama and above all else, passion, the PSA World Tour is the pinnacle of world squash.
 
About WSF
The World Squash Federation (WSF) is the global governing body for the sport of squash and celebrated its 50th anniversary in 2017.
The WSF is an International Federation of approaching 150 members and around 50,000 courts worldwide. There are also five Regional Federations under their auspices.
They are responsible for every facet of squash development and promotion, other than the professional tours which are overseen by the Professional Squash Association (PSA).

Impossible dream becomes reality as Hong Kong defeat Pakistan to become Asian squash champions for the first time

Max Lee leads team to glory over once-mighty Pakistan to complete double gold for the Hong Kong team who also lift the women’s title in Cheongju
 
An “impossible” dream became a reality for Hong Kong in Cheongju, South Korea on Sunday.
Hong Kong became Asian men’s team squash champions for the first time, ending a 32-year wait to mark an historical day for the city who celebrated double gold after also lifting the Asian women’s team title.
“This is a massive achievement. It proved that our bronze medal at the World Championships last year was no fluke,” said Hong Kong number one Max Lee Ho-yin after they defeated 15-time champions Pakistan on Sunday.
“We hope to capitalise on this success and continue to improve, with the Asian Games this summer as our next big target. This result is a huge confidence booster as we prepare for the Indonesia campaign. We did it in South Korea and now we feel we can do it again in Jakarta.”
Hong Kong first took part in the biennial Asian team championships in 1986 and have come agonisingly close to winning the men’s title on a number of occasions, including last year when they finished runners-up behind the once-mighty Pakistan team.
Buoyed by their third place finish at last year’s Worlds, Hong Kong, the number one seeds, knew they had the know-how and ability to overcome the Pakistanis, who entered the 15th edition of the tournament as the defending champions.
Lee and compatriot Leo Au Chun-ming both defeated their opponents in straight games, leaving the third member of the team, Yip Tsz-fung, the luxury of not having to play the dead rubber in the best-of-three tie.
Lee laid the foundation by defeating Farhan Zaman 3-0 (11-9, 11-9, 11-7) in the opening rubber, while Au dropped a game against Tayyab Aslam but gradually wore down his opponent as the Hong Kong number two prevailed 3-1 (11-6, 8-11, 11-0, 11-7) to complete the landmark win.
Hong Kong’s women’s team were only offered token resistance as they thrashed home team South Korea 2-0 in the final to lift their fourth title.
The Koreans had been cheered by their fans the previous day after they upset second seeds Japan to make their first appearance in the final, but the home fans were left in stunned silence after Hong Kong completed their demolition job against the hosts.
Hong Kong stalwarts Annie Au Wing-chi and Joey Chan Ho-ling, ranked 11th and 19th in the world respectively, proved too strong for their opponents to complete a fine campaign in South Korea.
Au won her opening tie against Ahn Eun-tschan with a quick-fire 11-4, 11-5, 11-5 victory, while Chan came back from one game down to beat Lee Ji-hyun 8-11, 11-0, 11-2, 11-7. Tong Tsz-wing was also spared from playing the dead rubber.


Nicol already Down Under as she eyes third straight gold

 Getting a feel: Nicol David going through the paces at the Oxenford Studios in preparation for the Commonwealth Games in Gold Coast.
Getting a feel: Nicol David going through the paces at the Oxenford Studios in preparation for the Commonwealth Games in Gold Coast.
PETALING JAYA: Malaysian squash queen Nicol David is totally committed to winning her third consecutive Common­wealth Games gold medal.
And the 34-year-old, who knows she has a daunting challenge ahead, has taken the initiative by checking into Gold Coast, Australia, early.
The eight-time world champion, who is coached by Aussie Liz Irving, has been getting a feel of the court at Oxenford Studios for the past week.
The squash competition is from April 5-15.
“I have been here since last week. My coach Liz is from Brisbane, so I thought this is the best opportunity to come a little early to prepare for the Games,” said Nicol.
“I just want to get a good feel of the court so that I can get off to a good start.”
The world No. 8 – who has going through a rough patch of form, including a first-round defeat at the Windy City Open in Chicago earlier this month – certainly needs a head start over an improving field of rivals.
The Penangite, who is the third seed, has a first-round bye before meeting either Colette Sultana of Malta or Cayman Islands’ Samantha Hennings in the second round.
She should progress to the last 16 to set up a meeting with Canada’s Samantha Cornett while England’s fifth seed Alison Waters is a potential quarter-final opponent.
If Nicol lives up to her seeding, she is then expected to meet the red-hot Joelle King of New Zealand in last four.
Besides Nicol, the other Malaysian women competing in the singles are national No. 2 S. Sivasangari and junior Aifa Azman.
Sivasangari is drawn to meet Scotland’s Alison Thomson after a first-round bye while Aifa will also face a Scottish player – 14th seed Lisa Aitken after a first-round bye.
In the men’s competition, the trio of Mohd Nafiizwan Adnan, Ng Eain Yow and Ivan Yuen will head Malaysia’s challenge in the singles.
The rest of the Commonwealth Games squad will only depart for Gold Coast on March 31.

CREDIT: The star online

We are all professionals and we realise we have to defend our Gold, says Dipika Pallikal


 no photo
Dipika Pallikal has admitted that although it is tough to imagine going to the CWG without coach Achraf El Karargui, the Squash team is determined to defend the Doubles’ Gold title. Pallikal is optimistic about her chances in Gold Coast as she aims for a medal in the Singles’ category as well.

India were rewarded with their first major medal in the sport of Squash four years ago at the 2014 Commonwealth Games in Glasgow when Dipika Pallikal and Joshna Chinappa beat the top-seeded Laura Massaro and Jenny Duncalf from England to clinch the Doubles’ Gold medal. A huge credit for the win went to coach Achraf El Karargui who helped the duo of Pallikal-Chinappa improve their game and clinch Gold for the country.
But all that has changed in the last four weeks, which seen Karagui step down from his role as coach of the Indian Squash contingent just weeks before the 2018 CWG in Gold Coast. Pallikal admitted that she was extremely ‘disheartened’ with the news but concedes that as professionals it is the team’s duty to defend their Gold in the upcoming tournament regardless of the circumstances.

"I cannot speak for others but it (not having a coach) is disheartening. I have been training with him for over two years and it is hard not to have him in your corner during the Games. Having said that, we are all professionals and we realise we have a very important job at hand, that is to defend our gold.
"He (Karargui) was the biggest asset for the team. I think he has changed a lot for Indian squash. For me, I know that I will be on a phone call with him everyday. It has all gone downhill in the last four weeks and now we are just trying to move forward," Pallikal told PTI.
Pallikal added that while things were different four years ago when the Indian’s were considered as underdogs at the tournament, it’s not the same case this time around. Pallikal has a first round bye in singles and the ninth seed is drawn to meet fifth seed Alison Waters of England in the Round of 16 in Gold Coast. Her compatriot, Eighth seed Chinappa is drawn to meet second seed Joelle King in the last eight stage.
"Our best chance is women's doubles but anything can happen in that format as we don't play it often. As of now, singles will be played first and we are focussed on that. There is a good chance of winning a singles medal too.
"Every match is going to be hard. Without a doubt it is going to be exhausting (playing three categories). But we have managed in the past, so hopefully this time it should not be much of an issue either," said Pallikal.
Pallikal added that her husband and cricketer, Dinesh Karthik, served as an inspiration for her ahead of the tournament. She opined that his match-winning six versus Bangladesh in the finals of the Nidahas Trophy was a reward for his hard work over the years.
"It has been very hard two years for both of us. We have hardly have spent anytime because of our hectic schedules. He totally deserves what he has achieved in the last one year and surely it gives me additional motivation to do well too," she said referring to Karthik, who cheered her on from the sidelines in Glasgow four years ago.

CREDIT: Sportcafe

Squash restructure aimed at simplifying tour and boosting earnings

 Egypt's world squash champions Raneem El Welily and Mohamed El Shorbagy
A new tour structure has been announced which aims to streamline professional squash and boost players' earnings.
The current Professional Squash Association (PSA) World Tour will be rebranded into two individual circuits, with qualification rounds scrapped.
A world tour will be complemented by a challenger tour, where selected tournament winners will earn a place at the 2018-19 World Championships.
The event in Chicago in early 2019 will have record $1m (£710,000) prize money.
That will be split equally across both the men's and women's draws and is up from $604,000 (£429,000) for this season's World Championships in Manchester.
The Championships will comprise of 56 PSA entrants and one wildcard, with the other seven places taken up by winners from the challenger tour in a 'Road to Chicago' competition. The women's tournament is expanding from 32 to 64 players.
Events on the world tour will offer prize money of $50,000 (£35,500) to $1m (£710,000) or above, with between $5,500 (£3,900) to $30,000 (£21,300) on offer at challenger events.
The new structure follows a partnership agreement last year between the PSA, which looks after the professional tour, and the sport's governing body, the World Squash Federation.
A satellite tour will be formed, which will see world, regional, national and junior tournaments integrated into the world rankings to offer a route into the professional circuit for up-and-coming players.
"We are committed to promoting a thriving and sustainable professional squash circuit and feel that simplifying the tour structure and easing the transition from junior squash to professional squash will help us to achieve those aims," said PSA Chief Executive Alex Gough.
Egypt's world squash champions Raneem El Welily and Mohamed El Shorbagy

Audio-based performance evaluation of squash players

Guys...ye all need to read this to see if you can digest and understand!

Audio-based performance evaluation of squash players

CREDIT:journals.plos...read complete Article Here

Abstract

In competitive sports it is often very hard to quantify the performance. A player to score or overtake may depend on only millesimal of seconds or millimeters. In racquet sports like tennis, table tennis and squash many events will occur in a short time duration, whose recording and analysis can help reveal the differences in performance. In this paper we show that it is possible to architect a framework that utilizes the characteristic sound patterns to precisely classify the types of and localize the positions of these events. From these basic information the shot types and the ball speed along the trajectories can be estimated. Comparing these estimates with the optimal speed and target the precision of the shot can be defined. The detailed shot statistics and precision information significantly enriches and improves data available today. Feeding them back to the players and the coaches facilitates to describe playing performance objectively and to improve strategy skills. The framework is implemented, its hardware and software components are installed and tested in a squash court.
 
IntroductionAt present in competitive sports there are a lot of talented sportsmen and the differences between individual performance are often very small to spot. It catalyses a race condition to be present already in the practising period, thus more and more coaches and players seek finding different means and aids to elaborate and make the preparation for the tournaments always more effective. There are a lot of new technological achievements available in the market. Small electronic devices are capable of measuring various metrics including those that are relevant for the sports, like heart rate and blood temperature and pressure registers, pedometers, speedometers and accelerometers to name a few. Using such devices is more than necessary since the results in a competition and then the final scores may depend on millesimal of millimeters. Another reason why to use measurement devices yielding objective performance metrics is because when sportsmen are overloaded in a performance, with adrenalin in their vein, it is hard if possible for them to spot and fix their failures. In certain types of sports a continuous or prompt feedback is definitely helpful, squash is one of them.
Squash is a very rapid ball and racquet game with typically 40-60 hit events per minute. Depending on the various surfaces the ball interacts during its flight defines the different shot classes. Some shot classes are very rare due to being tricky to deliver or may occur only in circumstances where the rally may seem already lost. So knowing the detailed statistics of various hits and shot patterns talks about the quality of the sportsmen and are very important information for both the coaches and the squash players. However, these data and their statistical analysis are not available at present because of the paste of squash. Given its fast speed the human processing of events enables the score registration in real-time only, but the recording of shot types and the detailed sequences of the shots are rendered definitely impossible. One possible solution might be to analyse videos of the matches using image processing as it has been shown to work for the tennis [1]. Though for the squash it turns out that this approach remains difficult even with the use of high speed and high resolution cameras, due to the small size of the ball and the view provided by the cameras. Traditionally cameras are placed behind the court, therefore the players will most often cover the sight of the ball during the match making the reconstruction of ball trajectories an inauspicious problem. To provide reliable statistics by this approach will require human processing and validation so in the end a thorough analysis of the tournament will cost many times of the duration of this sport events in man-hours.
In this study we introduce a framework to unhide these information based on the analysis of acoustic data. Playing squash produces characteristic sound patterns. The sound footprint of each rally is a projection of all the details about the strength and the position of the ball hitting various surfaces in the court. Naturally, this pattern, which maintains the natural order of the events, is contaminated by some additional noise. Recording the sound in more directions allows for inverting the problem and for giving statistical statements about where and what type of an event took place in the play. We are focusing on events generated by the ball hits, which serves as a basis for further analysis and the reconstruction of shot patterns or the ball trajectories. Note, the framework to be detailed can be applied to various other types of ball games.

Related work

Squash and soccer were the first sports to be analysed by ways of analysis systems. Formal scientific support for squash emerged at the late 1960s. The current applications of performance analysis techniques in squash are deeply investigated in the book of Stafford et al. [2].
One test that was developed by squash coach Geoffry Hunt is the “Hunt Squash Accuracy Test” (HSAT) [3], that is a reliable method used by coaches to assess shot hitting accuracy. The test is composed of 375 shots across 13 different types of squash strokes and it is evaluated based on a total score expressed as the number of successful shots.
Recent technological advances have facilitated the development of sport analytical software such as Dartfish video based motion analysis system [4, 5]. However, these systems still require a considerable amount of professional assistance.
To the best of our knowledge there is no previous research investigating the applicability of sound analysis techniques for squash performance analysis, therefore it is not possible to directly compare our system to existing solutions. In other application environments a wide literature can be found on real-time sound source localization that is the most closely related topic to our work. The emerging application of camera pointing in video conferencing environment motivated many research papers on the field of visual speaker localization [610]. A linear-correction least-squares estimation procedure is proposed in [68]. The simulation results in [7] show that the bias level of this technique is around 30 cm. In the work of Tobias Gehrig et al. [10] a method is presented to speaker tracking using audio-visual features, namely time delay of arrival estimation on microphone array signals and face detection on multiple camera images. The sound source localization is based on a maximum likelihood approach. In the experimental results the authors measure 57.2 cm root mean square error for the audio-only solution and 49.9 cm for the audio-video approach. One conceptually simple solution for source localization is beamforming [11], where the source location is estimated by calculating the steered output power of a beamformer over a set of candidate locations. Although this concept has advantages in speech localization and enhancement, it is computationally expensive and its resolution is too low for our purposes.

The measurement equipment

This study is based on the analysis of sound waves generated during the squash play. Among many other, squash is a game where various different sources of sound are present, including the players themselves (their sighing or their shoes squeaking on the floor), the ball hitting surfaces (like the walls, the floor or the racquet) and also external sources (including the ovation of the spectators or sound generated in an adjacent court). Here we focus on audio events related to the ball.
When planning the experiments the following constraints had to be investigated and satisfied. The framework should be fast in signal processing point of view, because the target information can be most valuable when in a competitive situation it helps fine tune tactical decisions made by the coach and/or the player. The cost of the equipment should be kept low and the installation of the sensors requires a careful design to prevent them from disrupting the play. As the spatial localization of the ball is one of the fundamental goals a lower bound to the sampling rate is enforced to remain able to differentiate between displaced sound sources.
In Fig 1 the hardware and software components are sketched. Hardware components include 6 audio sensors, three of which are omnidirectional microphones (Audio Technica ES945) sinking in the floor and the rest of them are cardioid microphones (Audio Technica PRO 45) hanging from the top. Amplification and sampling of the microphone signals are done by a single dedicated sound card (Presonus AudioBox 1818VSL) so that all channels in a sample frame are in synchrony. The highest sampling rate of the sound card is used (96 kHz), so by each new sample the front of a sound wave travels approximately 6 mm.
thumbnail

Fig 1. A schematic view of the components.
To process audio events in the squash court a three component architecture was designed.
According to their functionalities software components fall in the following groups. Signal processing is done in the analysis module, which include the detection of the audio events, the classification and the filtering of the detections and after matching event detections of more channels the localization of the sound source. While these steps of signal processing can be done real-time a storage module is also implemented so that the audio of important matches can be recorded. Recording of data helps training of the parameters of the classification algorithms, and it also enables a whole re-analysis of former data with different detectors and/or different classifiers. All output generated by the Analysis module is fed to the output queue. Hardware and software components are triggered and reconfigured via a web services API exposed by the Control interface. Finally, to be able to listen to what is going on in the remote court a Monitoring interface provides a mixed, downsampled and compressed live stream across the web.

The ball impact detection

The localization and the classification of ball hits both require the precise identification of the beginning of the corresponding events in the audio streams. The detection of ball impact events is carried out for each audio channels independently and in a parallel fashion, which speeds up the overall performance of the framework significantly. Different detection algorithms of various complexities were investigated two extreme cases are sketched here. The first model assumes that the background noise follows the normal distribution. An event is detected if new input samples deviate from the Gaussian distribution to a certain predefined threshold value. Next for each channels the mean and the variance estimates of a finite subset of the samples are continually updated according to the Welford’s algorithm [12].
The second method is an extension of the windowed Gaussian surprise detection by Schauerte and Stiefelhagen [13]. The algorithm tackles the problem evaluating the relative entropy [14]. It is first applied in the frequency domain and if there is a detection then a finer scale search is carried out in the time domain. The power spectrum of w-sized chunks of windowed data samples is calculated. Between detection regime the series of the power spectra is modelled by a w-dimensional Gaussian. The a priori parameters of the distribution are calculated for n elements in the past, and the posteriori parameters are approximated including the new power spectrum. The Kullback Leibler divergence between the a priori and the posteriori distributions exceeds a predefined threshold when a new detection takes place where primed parameters correspond to the posteriori distribution. The time resolution at this stage is w and to increase precision a new search is carried out in the time domain evaluating the Kullback Leibler divergence for 1-d data. In order to bootstrap a priori distribution parameters n samples from the former windows are used.

The localization of sound events

In this section we lay down a probabilistic model to determine the time and location of an audio event. For a unique event we denote these unknowns t and rev respectively. The location vector rev is a 3 dimensional array of Descartes coordinates (x, y, z), however, the calculation presented here also applies for lower dimension setups. The inputs required to find the audio event are the locations of the N + 1 detectors and the timestamps τi when these synchronized detectors sense the event (0 ≤ iN).
The probability that microphone i detects an event at (r, t) is where c is the speed of sound, ti = τit is the propagation delay and is the distance between the sound source and the microphone. The uncertainty σi depends on the characteristics of the microphone, which we will consider constant in the first approximation.
By introducing relative delays the joint probability of relative delays detected is The formula can be rearranged where is a quadratic function and in the expression for p the Gaussian integral follows The first order derivative f′ vanishes in , where is introduced for convenience.
After substitution of we arrive at This formula can be interpreted as a variance formula, which can be rewritten
A good approximation of the audio event maximizes the likelihood p, which at the same time minimizes , thus we seek the solution of equations.
In practice f behaves well and its minimum can be found by gradient descent method. Fig 2 shows a situation, where the ball hit the front wall and 6 microphones detect this event error free. To show the functions behaviour f is evaluated in the floor, in the front wall and in the right side wall. Finding the minimum of f takes less than ten gradient steps.
thumbnail

Fig 2. The visualization of the likelihood function.
The ball hit the front wall. can be evaluated in space given the positions of the sensors (marked by white disks) to find its minimum, which indicates where the event took place. (0.5 m from the right corner and 3 m above the floor, marked by a blue disk).
The likelihood based localization model is derived for a noiseless situation, assuming the perfect detection of samples in each channel. In real environment, however, noise is present and the error deviating the detection is exposed in the final result of the localization. In order to track this effect the method was numerically investigated as follows. 10000 points in the volume of the court is selected randomly and the sound propagation is calculated in each six microphones. Next for the ideal detections Gaussian noise is added in all channels, with increasing variation (σ = 1, 10, 50). In Fig 3 the noiseless case is compared to cases with increasing errors. In the figure the cumulative distribution of the error, ie. the difference between the randomly selected point and the location guess by the model is presented. Naturally, by increasing the detection error the error in the position guess is increasing, but the model performs very well, for poor signal detectors the error in localization is in the order of 10 cm.
thumbnail

Fig 3. The cumulative distribution of the localization error.
For a noiseless case most often localization will have an error comparable to the size of the ball. With a bad detector (σ = 50 samples) still the localization is exact in the order of 10 cm.

Classification

It is the task of the classification module to distinguish between the different sound events according to their origin. Sound events are classified based on the type of the surface that suffered from the impact of the ball. This surface can be the wall, the racquet, the floor or the glass. When the sound does not fit any of these classes, like the squeaking shoes, then it is classified as a false event. The classification enhances the overall performance of the system by two means. First, skipping to localize the false events speeds up the processing. And second, in doubtful situations when the calculated location of the event falls near to multiple possible surfaces, by knowing the type of the surface that suffered from the impact can reinforce the localization. For example a sound event localized a few centimetres above the floor could be generated by a racquet hit close to the floor or by the floor itself.
Classification utilizes feed-forward neural networks that had been trained with backpropagation [1518]. The training sets are composed of vectors belonging to 5461 audio events, which have been manually labelled. Based on these audio events two types of input were constructed for teaching.
In the first case temporal data is used directly. A vector element of the training set T1 is the sequence of the samples around the detections for each channels. where the channel index is dropped and d is a unique detection and w sets the length of the vector. Given the sampling rate 96 kHz and setting w = 300 the neural network is taught by 6.25 millisecond long data.
The second feature set T2 is built up of the power spectra. where denotes the discrete Fourier transform.
A single neural network model where all event classes are handled together performed poorly in our case. Therefore, separate discriminative neural network models were built for all four classes (racquet, wall, floor and glass impact) and for both of the training sets. It has also been investigated if any of the input channels introduce discrepancy. In order to discover this effect models were built and trained for each unique channels and another one handling the six channels together. Note, that not all possible combinations of the models were trained due to the fact that some channels poorly detected certain events, for example microphones near the front wall detected glass events very rarely.
In the training sets the class of interest was always under-represented. To balance the classifier the SMOTE [19] algorithm was used, which is a synthetic minority over-sampling technique. A new element is synthesized as follows. The difference between a feature vector from the positive class and one of its k nearest neighbours is computed. The difference is blown by a random number between 0 and 1, to be added to the original feature vector. This technique forces the minority class to become more general, and as a result, the class of interest becomes equally represented like the majority set in the training data.
Different network configurations were realized to find that for the direct temporal input a 20 hidden layer network (with 10 neurons in each layer) performed the best, while for the spectra input a 10 hidden layer (each layer with 10 neurons) is the best choice.

Analysis

In this section the performance of each modules of the framework and the datasets are presented.

Datasets

In order to analyse the components of the framework implementing the proposed methods two audio and video record sets were used. Datasets are available at https://figshare.com. Audio 1 was recorded on the 18th of May 2016 when a squash player was asked to target specific areas of the wall. This measurement was necessary to increase the cardinality of the different hits significantly in the training datasets T1 and T2, and it was also manually processed to be able to validate the operations of the detector and the localization components. Audio 2 resembles data in a real situation as it contains a seven minutes squash match recorded on the 8th of March 2016. Table 1 summarizes the details of these audio recordings.
thumbnail

Table 1. The content of the audio files.
Training the neural network models require properly labelled datasets. After applying the ball impact detection algorithm to the audio records the timestamps of the detected events were manually categorized as front wall event, racquet event, floor event or glass event. In the categorization procedure video files helped in doubtful situations. Every sudden sound effect that was detected by the algorithm but does not belong to these relevant classes was labeled as false event. In Audio 1 prescribed audio events were generated and recorded and it does not contain any false events. In contrast, Audio 2 was captured in a real situation and it presents several false events by nature.

Detection results

The performance of the detector is analysed by comparing the timestamp reported by the detector ddetector and the human readings dhuman. For Audio 1 in Fig 4 the cumulative probability distribution of the time difference is shown for each channel and in Table 2 the average error and its variance are shown grouped by the event types present in the dataset. One can observe that the detectors in channels ch4 and ch5 perform poorly for front wall and racquet events. When estimating the position discarding one of or both of these channels will enhance the precision of the localization. However, for floor events, these two channels performed the best along with channel ch2. For glass events the smallest deviations were measured on channels ch2, ch3 and ch5.
thumbnail

Fig 4. The error of the detector.
The detection error is defined as the difference between the timestamps generated by the module and read by a human.
thumbnail

Table 2. The class and channelwise error of the detector.
In Table 3 the error statistics for dataset Audio 2 is shown. Intensive events, like front wall impacts, can be detected precisely, whereas the detection of milder sounds like a floor or glass impact is less accurate.
thumbnail

Table 3. Classwise error of the detector.
The false discovery and the false negative rate of the detector were examined on Audio 2. False positives are counted if detector signals for a false event, and false negatives are the missing detections. The results are summarised in Table 4.
thumbnail

Table 4. Performance of the detector.

Classification results

Approaching the problem at first and to use as much information as possible to teach the neural networks a large training set was constructed of the union of the detections of all the six channels. However, this technique gave poorer results than treating all the channels separately. The different settings of the microphones and the distinct acoustic properties of the squash court at the microphone positions are found to be the reasons of that phenomenon.
Eight-fold cross-validation [20] was used on the datasets to evaluate the performance of the classifiers. Three measures are investigated closer: the accuracy, the precision and the recall. Accuracy (in Fig 5) is the ratio of correct classifications and the total number of cases examined (). Precision (in Fig 6) is the fraction constrained to the relevant cases (). Recall (in Fig 7) is the fraction of relevant instances that are retrieved ().
thumbnail

Fig 5. The classifiers’ accuracy.
The classwise accuracy of each channel is presented in T1 (blue) and T2 (red) input sets. Front wall classification gives high accuracy on all channels in both sets. It is interesting to observe that floor classification is more accurate in input T2. Racquet classification performs best on channel 2 in both sets.
thumbnail

Fig 6. The classifiers’ precision.
The classwise precision of each channel is presented in T1 (blue) and T2 (red) input sets. Front wall classification gives high precision in input T1. The precision of floor classification is low. Racquet classification still performs best on channel 2. The precision of glass classification is only acceptable on channel 4.
thumbnail

Fig 7. The classifiers’ recall.
The classwise recall of each channel is presented in T1 (blue) and T2 (red) input sets. The performance of front wall classification is reliable. The recall of racquet classification is high on channels 1 and 2 in both sets. However, the performance of floor and glass classifications is low.
Table 5 summarises the results of the best classifiers for each class. It can be seen that the classification of the front wall and the racquet events is reliable. However, the precision and the recall of floor and glass events are poor. The reason for it is that these classes are under-represented in the data sets. Whenever x, an unseen sample comes, the best classifiers of each class are applied on the new element. The prediction of the class label to which x belongs to is computed by the following formula: where C is the set of class labels without the class of false events and fk(x), cutk and preck are the confidence, the cutoff value and the precision of the best classifier in class k respectively.
thumbnail

Table 5. The classwise performance of the best classifiers.
Fig 8 depicts the combined output generated by the detector and the classifier modules. A 1.77 seconds long segment of channel 1 audio samples are grabbed from Audio 2. Detections and resolved classes are also shown. From the snapshot one can observe the different intensities of the events. Generally the change in the ball’s moment happens when a racquet or a front wall impacts and the sample amplitudes are higher, whereas floor and glass events tend to generate lower intensity and are harder to detect.
thumbnail

Fig 8. Labelled audio signal.
1.77 second long samples from channel ch1 in Audio 2. Detected timestamps and the event classes are marked.

Localization results

Based on the geometry of the court, the placement of the microphones and using the localization technique detailed in this study for each set of detection timestamps the 3-d position of the source of the event can be estimated. In case not all source channels provide a detection of the event localization is still possible. Four or more corresponding timestamps will yield a 3-d estimate, whereas with three timestamps the localization of events constrained on a surface (e.g. planes like wall or floor) remains possible.
In Fig 9 the located events present in dataset Audio 1 are shown. In this measurement scenario the player was asked to hit different target areas on the front wall. It was a rapid exercise, as the ball was shot back at once. Only a few times the ball hit the floor, most of the sound is composed of alternating racquet and front wall events. In Fig 10 the front wall events are shown. The target areas can be seen clearly, and also it is visible the spots scatter a little more on the left. The reason could be the player being right handed or the fact the target area was hit later during the experiment and the player showed tiredness.
thumbnail

Fig 9. The position of impacts.
Visualize the localized events embedded in 3-d.
thumbnail

Fig 10. Front wall impacts.
Gray squares embrace the eight target areas.
Measuring the error of the localization method is not straight forward because the ball hitting the main wall does not leave a mark, where the impact happened and there was no means to take pictures of these events. Taking advantage of the geometry of the front wall an error metric can be defined for front wall events. The error δ is defined by the offset of the approximated location from the plane of the front wall. In Fig 11 the error histogram is shown. The mean of δ should vanish and the smaller its variance the better the framework located the events. From this exercise one can read the standard deviation is σ(δ) < 3 cm, which is smaller then the size of the squash ball.
thumbnail

Fig 11. The front wall offsets.
The distribution of the offset δ from the front wall (σ(δ) ≈ 0.02 m).
Another way to define the error is based on relying on human readings of the events. In the dataset Audio 1 all of the sound events were marked by human as well as by the detector algorithm. Localizing the events using both inputs the direct position difference can be investigated. The mean difference between the positions is 11.8 cm and their standard deviation is 39.9 cm.

Discussion

Our results support that in sports, where the relevant sound patterns are distinguishable, careful signal processing allows the localisation of shots. The described system is optimized for handling events and as a consequences the real-time analysis of data is possible, which is important to give an instant feedback. The framework can be extended to provide higher level statistics of events such as the evolution of shots types. From the wide range of possible applications we highlight three use cases. Firstly, during a match the players can get to know their precision in short time and if is necessary they can change their strategy. Secondly, during practice coaches can track the development of the players hit accuracy. Or thirdly, certain exercises can be defined, which can be automatically and objectively evaluated, without the need for the coach be present during the exercise.
In this study our framework was adopted to squash. In theory it can be extended to any other sports where the stereotypical events are associated with a specific sound pattern. In those applications, where typical patterns are present but the surroundings introduce significant amount of noise, a solution could be to use additional microphones with possibly special characteristcs to record the noise allowing to subtract its contribution from all other input signals. For example in tennis games played in the open field.

Ethics statement

In this study human participants were instructed to carry out specific squash excercises. Participants were informed beforehands about the fact that during the excercise the sound is to be recorded. During the excercises the sound emerging mainly from the ball impacts was recorded. The recording itself do not contain any sensitive information. Along with the raw recordings no additional information about the participant is saved or published. Participants do not object that these recordings are made public.

Acknowledgments

The hardware components enabling this study are installed at Gold Center’s squash court. We thank them for this opportunity and squash coach Shakeel Khan for the fruitful discussions.
Authors thank the Hungarian National Research, Development and Innovation Office under Grant No. 125280 and the grant of Ericsson Ltd.

Egypt & England to contest Women’s World Final in Dalian

For the fifth year time in seven years, Egypt and England will contest the final of the CGG WSF Women’s World Team Squash Championship in...