Man
Professional
- Messages
- 3,059
- Reaction score
- 585
- Points
- 113
We explore the latest method for detecting hidden cameras and microphones, described in one of the most interesting works on personal security this year.
In a 2019 survey of travelers who used short-term rental services, 11% of study participants admitted to finding CCTV cameras in their rental accommodation. In addition, about two-thirds of respondents were concerned that unscrupulous apartment owners might use hidden cameras.
We recently wrote about how Apple AirTags are used to stalk people and steal cars. The problem of hidden filming in rented accommodation has the same roots - seemingly useful devices are used for harm.
It is unlikely that the problem of hidden surveillance will go away anytime soon - rather the opposite. Firstly, a simple wireless video camera that transmits data in real time via Wi-Fi can be purchased for pennies. Secondly, at the current stage of technological development, even cheap surveillance devices have become quite miniature: they can be easily hidden in the interior.
There are several methods for finding hidden cameras. The simplest of them is to scan the room in the dark with a phone camera to detect video cameras with infrared illumination. However, the effectiveness of known methods leaves much to be desired, and the demand for more reliable methods for detecting surveillance devices is growing along with their increasing prevalence.
Therefore, attempts are made from time to time to create more high-tech anti-spy technologies. The most advanced system was recently proposed by researchers from Carnegie Mellon University in the USA. They found a way that in the future (if the idea is developed) will allow to detect hidden IoT devices in an unfamiliar room using an application on a smartphone.
Then the guests walk around the perimeter of all the rooms with their phone. This is necessary so that the smartphone can determine the coordinates of the detected transmitters. As a result, the application not only shows a list of detected "smart" devices, but also allows you to determine their location. IoT devices are displayed in "augmented reality" mode: information about the location of smart devices is superimposed on the image from the phone camera. The researchers recorded a demonstration of the application on video:
It looks nice, but how does it work? The authors of the study assume that the typical landlord renting out an apartment through AirBnb or a similar service will not waste time installing truly sophisticated spy equipment - it is expensive and, in the vast majority of cases, unnecessary. Instead, the landlord is more likely to buy something simple and cheap in the store.
It’s not that hard to pinpoint a rough list of devices that are potentially dangerous to privacy. The most obvious choice is a video camera that transmits image and sound via Wi-Fi. It could also be a smart speaker set up to constantly “listen in” to what’s going on in the room, or a smart TV. Or even a specialized, but “civilian” security system, such as Amazon Ring.
All Wi-Fi-connected devices operate in a common frequency range and must coordinate data transfer with each other. Based on this open information, the Lumos system can identify nearby wireless devices, distinguish a smart bulb from a webcam, and even indicate an approximate location with an accuracy of one and a half meters.
The spy device detection system adds an AR hint to the image from the smartphone camera about the presence of a wireless camera nearby.
The thing is, for the system to work, it is not enough for it to connect to an access point and communicate only with it. Instead, Lumos scans all data packets transmitted in the air from all devices around. But in all smartphones, the ability to "listen to the air" is blocked.
Theoretically, this could be done on a “rooted” Android device (i.e., a gadget where the owner has received superuser rights, giving full access to the system), but the researchers took a different approach. They built two test systems: the first consisted of a smartphone and a Raspberry Pi microcomputer connected to it, and the second was a laptop with a special camera.
In the first case, the Raspberry Pi was listening to the Wi-Fi broadcast, in the second case, the laptop itself (it is easier to implement such functionality on a computer). Accordingly, either a smartphone camera or an augmented reality camera connected to a laptop was used to visually detect hidden devices.
How to find hidden Wi-Fi devices in unfamiliar surroundings using your phone.
Next, the researchers applied technology to identify the type of devices based on their behavior. A simple example: a smart switch might access a server once a minute and not show itself in any other way, but a webcam transmits a lot of data and does so constantly. Using more than 50 such characteristics and machine learning, the authors of the work created “digital fingerprints” of typical IoT devices.
As a result, there is no need to keep a database of suspicious devices in the application and constantly update it - the Lumos system will, with a high probability, determine that you have a Wi-Fi camera in your room (even if it has never encountered this particular model before) based on its characteristic behavior when transmitting data .
Another example from the Lumos demo: a smart speaker found with its help.
The overlay of the approximate coordinates of the found devices on the picture from the phone camera was made on an Apple smartphone using the standard ARkit library. This library uses various sensors of the phone to create a three-dimensional model of the room. At the same time, the location of hidden connected devices is determined.
To do this, the researchers used another feature of the Wi-Fi module - it constantly measures the signal strength from both the nearest access point and other wireless modules. Thus, if you walk around the room, measure the signal level at different points and do some simple calculations, you can determine the position of the hidden device in space.
Map of indoor Wi-Fi device search by signal strength.
However, the results that the researchers managed to achieve still inspire optimism:
This means that such a system can really help detect unpleasant surprises in rented accommodation or a hotel. However, there are limitations. We have already mentioned above that it is impossible for a regular smartphone to scan all Wi-Fi traffic. It is unlikely that manufacturers will quickly agree to change this.
In addition, there are now a huge number of Wi-Fi devices in every home, and radio waves pass through walls with some success, especially if they are not reinforced concrete. Because of this, you can accidentally discover a camera that is installed at your neighbors or even located on the outer wall of the house - and why accuse the landlord of surveillance in vain.
Oh yeah, attempts to detect a hidden camera are completely wasted if the intruder connected it via a wired connection or prefers offline viewing of the footage, recording the video the old-fashioned way on a flash drive.
Because of these two and a half considerations, we are unlikely to see a commercial implementation of the Lumos system in the foreseeable future. It turned out to be too "hacker" (in a good sense). Any commercial company that undertakes the development will face a lot of problems and complaints. And both from clients (they didn't find the camera, but it was there), and from property owners (there was no camera, but the program shows the opposite).
If it works, it will most likely be a thing for tech enthusiasts. It might even require buying a specially modified smartphone. However, in such a “complex” scenario, the system could well be free (apart from the price of this special smartphone) — the authors of the work have promised to make the source code for the prototypes freely available.
Source
In a 2019 survey of travelers who used short-term rental services, 11% of study participants admitted to finding CCTV cameras in their rental accommodation. In addition, about two-thirds of respondents were concerned that unscrupulous apartment owners might use hidden cameras.
We recently wrote about how Apple AirTags are used to stalk people and steal cars. The problem of hidden filming in rented accommodation has the same roots - seemingly useful devices are used for harm.
It is unlikely that the problem of hidden surveillance will go away anytime soon - rather the opposite. Firstly, a simple wireless video camera that transmits data in real time via Wi-Fi can be purchased for pennies. Secondly, at the current stage of technological development, even cheap surveillance devices have become quite miniature: they can be easily hidden in the interior.
There are several methods for finding hidden cameras. The simplest of them is to scan the room in the dark with a phone camera to detect video cameras with infrared illumination. However, the effectiveness of known methods leaves much to be desired, and the demand for more reliable methods for detecting surveillance devices is growing along with their increasing prevalence.
Therefore, attempts are made from time to time to create more high-tech anti-spy technologies. The most advanced system was recently proposed by researchers from Carnegie Mellon University in the USA. They found a way that in the future (if the idea is developed) will allow to detect hidden IoT devices in an unfamiliar room using an application on a smartphone.
Detect spy devices with the app
The authors of the work decided to implement the following scenario. "Test guests" receive keys to a rented apartment. After checking in, they launch a special Lumos application on their smartphone and wait for about half an hour.Then the guests walk around the perimeter of all the rooms with their phone. This is necessary so that the smartphone can determine the coordinates of the detected transmitters. As a result, the application not only shows a list of detected "smart" devices, but also allows you to determine their location. IoT devices are displayed in "augmented reality" mode: information about the location of smart devices is superimposed on the image from the phone camera. The researchers recorded a demonstration of the application on video:
It looks nice, but how does it work? The authors of the study assume that the typical landlord renting out an apartment through AirBnb or a similar service will not waste time installing truly sophisticated spy equipment - it is expensive and, in the vast majority of cases, unnecessary. Instead, the landlord is more likely to buy something simple and cheap in the store.
It’s not that hard to pinpoint a rough list of devices that are potentially dangerous to privacy. The most obvious choice is a video camera that transmits image and sound via Wi-Fi. It could also be a smart speaker set up to constantly “listen in” to what’s going on in the room, or a smart TV. Or even a specialized, but “civilian” security system, such as Amazon Ring.
All Wi-Fi-connected devices operate in a common frequency range and must coordinate data transfer with each other. Based on this open information, the Lumos system can identify nearby wireless devices, distinguish a smart bulb from a webcam, and even indicate an approximate location with an accuracy of one and a half meters.
The spy device detection system adds an AR hint to the image from the smartphone camera about the presence of a wireless camera nearby.
Lumos in detail
Many will say: enough details, where can I download the program? Unfortunately, nowhere yet. Moreover, in working prototypes it was not possible to implement all the functionality using only smartphones. And all because the needs of the Lumos system somewhat exceed their typical capabilities in terms of Wi-Fi.The thing is, for the system to work, it is not enough for it to connect to an access point and communicate only with it. Instead, Lumos scans all data packets transmitted in the air from all devices around. But in all smartphones, the ability to "listen to the air" is blocked.
Theoretically, this could be done on a “rooted” Android device (i.e., a gadget where the owner has received superuser rights, giving full access to the system), but the researchers took a different approach. They built two test systems: the first consisted of a smartphone and a Raspberry Pi microcomputer connected to it, and the second was a laptop with a special camera.
In the first case, the Raspberry Pi was listening to the Wi-Fi broadcast, in the second case, the laptop itself (it is easier to implement such functionality on a computer). Accordingly, either a smartphone camera or an augmented reality camera connected to a laptop was used to visually detect hidden devices.
How to find hidden Wi-Fi devices in unfamiliar surroundings using your phone.
Next, the researchers applied technology to identify the type of devices based on their behavior. A simple example: a smart switch might access a server once a minute and not show itself in any other way, but a webcam transmits a lot of data and does so constantly. Using more than 50 such characteristics and machine learning, the authors of the work created “digital fingerprints” of typical IoT devices.
As a result, there is no need to keep a database of suspicious devices in the application and constantly update it - the Lumos system will, with a high probability, determine that you have a Wi-Fi camera in your room (even if it has never encountered this particular model before) based on its characteristic behavior when transmitting data .
Another example from the Lumos demo: a smart speaker found with its help.
Search for IoT devices in augmented reality
The most spectacular part of the study is how the image from the smartphone camera is superimposed with cues about the location of nearby IoT devices. Lumos uses augmented reality (AR) technology, that is, it embeds virtual objects into the image of the real environment transmitted to the phone’s display from the camera.The overlay of the approximate coordinates of the found devices on the picture from the phone camera was made on an Apple smartphone using the standard ARkit library. This library uses various sensors of the phone to create a three-dimensional model of the room. At the same time, the location of hidden connected devices is determined.
To do this, the researchers used another feature of the Wi-Fi module - it constantly measures the signal strength from both the nearest access point and other wireless modules. Thus, if you walk around the room, measure the signal level at different points and do some simple calculations, you can determine the position of the hidden device in space.
Map of indoor Wi-Fi device search by signal strength.
Difficulties of implementation
Here we must remind once again that we are talking about scientific research, not about developing a commercial product. The authors do not have the task of selling us anything at all. It is even more interesting that they were able to assemble a full-fledged tracking device detection system, and did not show only some part of it, leaving the reader to figure out the rest.However, the results that the researchers managed to achieve still inspire optimism:
- The accuracy of determining the type of devices was 95-98%. The chance of error is quite small.
- The location of IoT devices is determined with an accuracy of up to one and a half meters - this is quite enough to later find the hidden camera visually.
- The time required to detect nearby IoT devices is 30 minutes. Of these, the smartphone simply lies there for 27 minutes and collects data packets transmitted by other devices. The user will need to walk around the room for another three minutes to determine the coordinates of the detected transmitters.
This means that such a system can really help detect unpleasant surprises in rented accommodation or a hotel. However, there are limitations. We have already mentioned above that it is impossible for a regular smartphone to scan all Wi-Fi traffic. It is unlikely that manufacturers will quickly agree to change this.
In addition, there are now a huge number of Wi-Fi devices in every home, and radio waves pass through walls with some success, especially if they are not reinforced concrete. Because of this, you can accidentally discover a camera that is installed at your neighbors or even located on the outer wall of the house - and why accuse the landlord of surveillance in vain.
Oh yeah, attempts to detect a hidden camera are completely wasted if the intruder connected it via a wired connection or prefers offline viewing of the footage, recording the video the old-fashioned way on a flash drive.
Because of these two and a half considerations, we are unlikely to see a commercial implementation of the Lumos system in the foreseeable future. It turned out to be too "hacker" (in a good sense). Any commercial company that undertakes the development will face a lot of problems and complaints. And both from clients (they didn't find the camera, but it was there), and from property owners (there was no camera, but the program shows the opposite).
If it works, it will most likely be a thing for tech enthusiasts. It might even require buying a specially modified smartphone. However, in such a “complex” scenario, the system could well be free (apart from the price of this special smartphone) — the authors of the work have promised to make the source code for the prototypes freely available.
Source
