Abstract: Audio-based event detection poses a number of different challenges that are not encountered in other fields, such as image detection. Challenges such as ambient noise, low Signal-to-Noise Ratio (SNR) and microphone distance are not yet fully understood. If the multimodal approaches are to become better in a range of fields of interest, audio analysis will have to play an integral part. Event recognition in autonomous vehicles (AVs) is such a field at a nascent stage that can especially leverage solely on audio or can be part of the multimodal approach. In this manuscript, an extensive analysis focused on the comparison of different magnitude representations of the raw audio is presented. The data on which the analysis is carried out is part of the publicly available MIVIA Audio Events dataset. Single channel Short-Time Fourier Transform (STFT), mel-scale and Mel-Frequency Cepstral Coefficients (MFCCs) spectrogram representations are used. Furthermore, aggregation methods of the aforementioned spectrogram representations are examined; the feature concatenation compared to the stacking of features as separate channels. The effect of the SNR on recognition accuracy and the generalization of the proposed methods on datasets that were both seen and not seen during training are studied and reported.
Audio-Based Event Detection at Different SNR Settings Using Two-Dimensional Spectrogram Magnitude Representations Ioannis Papadimitriou, Anastasios Vafeiadis, Antonios Lalas, Konstantinos Votis, Dimitrios Tzovaras Journal: Electronics Date: 29 September 2020, DOI: 10.3390/electronics9101593 Download:Publisher’s version (Gold Open Access)
https://h2020-avenue.eu/wp-content/uploads/2021/06/mdpi-logo.png3311000Dimitri Konstantashttp://avenue.unige.ch/wp-content/uploads/2019/04/avenue-logo-1.pngDimitri Konstantas2020-09-29 20:28:092021-06-04 12:55:59New Scientific Article: Audio-Based Event Detection at Different SNR Settings Using Two-Dimensional Spectrogram Magnitude Representations
Source: SWI Swissinfo Author: Simon Bradley Date: September 24, 2020
Over the past five years, various kinds of autonomous vehicles, including buses, have popped up on Swiss roads. But though testing continues, a driverless future might yet be some years away.
On a crisp autumn morning in the Geneva
countryside, a bright orange-and-white electric bus is winding through
the leafy 36-hectare grounds of the Belle-Idée hospital site.
The
bus is trundling along a gravel path when suddenly a patient and a
nurse step out from behind a tree. The vehicle brakes sharply, a bell
rings out, a “keep your distance” sign flashes at the front and rear.
The couple steps back, and the bus continues slowly on its way.
The
toy-like shuttle – empty apart from a safety operator and guided by
sensors, GPS and radar – is the centrepiece of a unique driverless
public transport experiment.
“It’s a world first for a public transport service,” says Dimitri Konstantas, director of the Information Science Institute at the University of Geneva, who is coordinating the project. “Most sites and lines have a fixed route… but here the difference is that there is no route. You can go anywhere.”
Testing, testing
This
summer, a small team began testing the ten-seater vehicles, mapping out
the huge Belle-Idée park and its obstacles. In parallel, a Geneva-based
start-up, MobileThinking, has been putting the finishing touches to an
app that will be tried out by the first passengers before the end of the
year.
Then when the project is up and
running in the next couple of years, patients, visitors and staff will
be able to get around the sprawling complex by using their smartphone to
order one of three buses offering an on-demand door-to-door service.
External Content
Sign up! Stay connected to Swiss science news and trends
Users will be able to locate a bus via the app,
then send a pick-up request. Software from the Lausanne firm
Bestmile will indicate when a bus is available and what the journey time
will be. A fleet management system will then adapt the vehicle’s route
according to other passenger requests.
The
idea is to have a completely automated system with a safety operator
back at a central depot monitoring the vehicles, says Jeroen Beukers, an
autonomous vehicle expert who is running the project for the Geneva
public transport authorities (TPG).
“Next
week we are installing electric doors on the bus depot. In the future,
you’ll make a booking on your phone, the depot doors will open
automatically, a charged bus will pick you up from A and take you to B
and then either return to the depot or continue onwards to pick up
someone else,” he says.
European project
The
project is not Geneva’s first driverless bus trial: TPG has been
successfully running an automated shuttle on a fixed circular route in
Meyrin since 2018.
The Belle-Idée project was selected as part of a four-year European driverless vehicle initiative known as AVENUE (Autonomous
Vehicles to Evolve to a New Urban Experience). The European consortium,
funded by the European Union’s Horizon 2020 programme, includes pilot
schemes in Lyon (France), Luxembourg and Copenhagen (Denmark).
Driverless vehicles in Switzerland
Over the past five years a growing number of other Swiss cities and transport companies have experimented
with driverless vehicles on fixed routes (see infobox). This includes
Sion, the capital of canton Valais, which in 2016 became the first Swiss
city to launch an autonomous bus service in collaboration with Swiss Post.
Apart
from the odd minor accident, the results of these trials have been
generally positive, with thousands of passengers now regularly taking
driverless shuttles.
Swiss ‘pioneer’
The
trials have allowed Switzerland “to position itself as a pioneer” in
this field, according to Marina Kaempf, spokesperson for the Federal
Roads Office.
In most cases the tests were
well accepted by the public, with municipalities and cantons developing
“realistic” projects to show what the vehicles can do, she tells
swissinfo.ch.
But current technologies still don’t allow
vehicles that are 100% driverless – i.e. without a safety operator – to
be used commercially, the Roads Office says. The exchange of data
between driverless vehicles and the outside environment also needs to be
improved.
“Longer term, you can imagine
driverless buses running more permanently on certain lines when their
technologies have been perfected,” says Kaempf. But in the short-term,
while autonomous pilot schemes will continue in their current form, the
Roads Office is not planning to increase them or turn them into
commercial ventures.
Meanwhile, in parallel
to these trials, the government is gearing up for the wider use of
driverless vehicles on Swiss roads in the coming years. In August, it
launched a consultation process
to revise the Federal Law on Road Traffic. Part of the proposal aims to
improve the legal basis for automated driving and future testing, and
to ensure Switzerland can adapt to any international developments in
this field.
Huge challenges ahead
Despite Switzerland pushing ahead, not all mobility experts are convinced by driverless buses.
“Maybe
they’ll have made big progress in 20 years, but at the moment
autonomous buses are a bit of a gadget,” says Vincent Kaufmann, a
professor at the Federal Institute of Technology Lausanne (EPFL) and
scientific director of the Mobile Lives Forum in Paris.
“What’s
interesting is not so much driverless buses, but shared autonomous
vehicles, like taxis. We’ll continue to need trams, trains and buses,
where you carry 100-200 passengers. But if the shared autonomous vehicle
can replace the individual car in the city that’ll be a real gain.”
Driverless challenges – both regulatory and
technical – remain huge, raising many questions. Will the introduction
of autonomous buses, taxis and privately-owned driverless vehicles just clog up the roads if they are not correctly regulated?
How safe will such vehicles be? Will autonomous buses be used downtown
or just in suburban areas? How will private data used by autonomous
vehicles and passengers be protected? What legal responsibility will
public transport firms and drivers have for their autonomous vehicles?
At
the technical level, Konstantas feels driverless vehicles still have a
long way to go before they can correctly identify objects and anticipate
people’s behaviour in the streets.
“Tesla is working on it, but I doubt they’ll be able to do that within 15-20 years,” he says.
He
also sees data protection as a big issue. “We’re not allowed to use the
data of people walking around in order to learn from it,” he says. “Our
system is programmed. It’s not dynamic learning or AI – we don’t have
that yet.”
“What we’re doing here is experimental. Is it possible
to build the future? We don’t know. Is this going to be useful or not?
We don’t know. But we’re going to try.”
On a crisp autumn morning in the Geneva
countryside, a bright orange-and-white electric bus is winding through
the leafy 36-hectare grounds of the Belle-Idée hospital site.
The
bus is trundling along a gravel path when suddenly a patient and a
nurse step out from behind a tree. The vehicle brakes sharply, a bell
rings out, a “keep your distance” sign flashes at the front and rear.
The couple steps back, and the bus continues slowly on its way.
The
toy-like shuttle – empty apart from a safety operator and guided by
sensors, GPS and radar – is the centrepiece of a unique driverless
public transport experiment.
“It’s a world first for a public transport service,” says Dimitri Konstantas, director of the Information Science Institute at the University of Geneva, who is coordinating the project. “Most sites and lines have a fixed route… but here the difference is that there is no route. You can go anywhere.”
https://h2020-avenue.eu/wp-content/uploads/2020/09/SWI_swissinfo.ch_Logo_2018.svg_.png2421280Dimitri Konstantashttp://avenue.unige.ch/wp-content/uploads/2019/04/avenue-logo-1.pngDimitri Konstantas2020-09-24 11:00:302021-06-01 20:33:31TV/Newspaper: SWI Swissinfo - The slow but steady progress of driverless buses in Switzerland
Abstract: Autonomous public transport is not a fantasy anymore. At least 92 experimentations with autonomous shuttles for collective transport (ASCTs) have been running in different cities around the world. The deployment of autonomous shuttles opens up a wide range of possibilities in rethinking urban mobility, with a greater focus on users, new services, and reduced urban traffic congestion. The prospective analysis of these possibilities produces original, innovative forecasts with new scenarios for ASPT based on three strong drivers of change, i.e., technological innovation, new mobility behavior, and new business models for public transport. Furthermore, the deployment of ASCTs will considerably modify costs of urban externalities and extend inclusion. The chapter presents how the implementation of autonomous shuttles may revolutionize the public transport.
Autonomous Vehicles toward a Revolution in Collective Transport Sylvie Mira Bonnardel, Fabio Antonialli and Danielle Attias Book: Autonomous Vehicle and Smart Traffic Date: 9 September 2020, DOI:10.5772/intechopen.89941 Download:Publisher’s version (Gold Open Access)
https://h2020-avenue.eu/wp-content/uploads/2021/06/intechopen.png3001000Dimitri Konstantashttp://avenue.unige.ch/wp-content/uploads/2019/04/avenue-logo-1.pngDimitri Konstantas2020-09-09 20:42:482021-06-04 12:58:54New Scientific Article: Autonomous Vehicles toward a Revolution in Collective Transport
Abstract: Autonomous vehicles (AVs) are already operating on the streets of many countries around the globe. Contemporary concerns about AVs do not relate to the implementation of fundamental technologies, as they are already in use, but are rather increasingly centered on the way that such technologies will affect emerging transportation systems, our social environment, and the people living inside it. Many concerns also focus on whether such systems should be fully automated or still be partially controlled by humans. This work aims to address the new reality that is formed in autonomous shuttles mobility infrastructures as a result of the absence of the bus driver and the increased threat from terrorism in European cities. Typically, drivers are trained to handle incidents of passengers’ abnormal behavior, incidents of petty crimes, and other abnormal events, according to standard procedures adopted by the transport operator. Surveillance using camera sensors as well as smart software in the bus will maximize the feeling and the actual level of security. In this paper, an online, end-to-end solution is introduced based on deep learning techniques for the timely, accurate, robust, and automatic detection of various petty crime types. The proposed system can identify abnormal passenger behavior such as vandalism and accidents but can also enhance passenger security via petty crimes detection such as aggression, bag-snatching, and vandalism. The solution achieves excellent results across different use cases and environmental conditions.
Real-Time Abnormal Event Detection for Enhanced Security in Autonomous Shuttles Mobility Infrastructures Dimitris Tsiktsiris, Nikolaos Dimitriou, Antonios Lalas, Minas Dasygenis, Konstantinos Votis, Dimitrios Tzovaras Journal: Sensors Date: 1 September 2020, DOI: 10.3390/s20174943 Download: Publisher’s version (Gold Open Access)
https://h2020-avenue.eu/wp-content/uploads/2021/06/mdpi-logo.png3311000Dimitri Konstantashttp://avenue.unige.ch/wp-content/uploads/2019/04/avenue-logo-1.pngDimitri Konstantas2020-09-01 20:36:342021-06-04 12:57:43New Scientific Article: Real-Time Abnormal Event Detection for Enhanced Security in Autonomous Shuttles Mobility Infrastructures