Sunday 22 November 2009

Audio Forensic with Cedar

A few weeks ago, I joined the Audio Forensic Training which was jointly conducted between Forensic Laboratory Centre of Indonesian National Police Headquarters and Cedar Cambridge. In this training, we developed the latest techniques on noise filtering by using Cedar instrument which was installed in the Audio Laboratory at my office. According to Dr. David Robinson who was also the instructor at this training, the audio lab we have is the best Cedar lab in South East Asia.

In this training, we praticed to remove a wide range of noise. In some cases, the voice recorded is not clear because of the noise, even it can not be heard at all. The noise sounds are much louder than the human voice. With the assisstance of Cedar providing many powerful filtering modules, we were successful to remove the noise and to make the human voice to be clear to listen to. Besides this, Cedar also provides feature to recognise the editing line in the case of when the voice recording is edited. Cedar can detect the time when the editing occured by displaying a vertical line. Through this line, we can know there is a change before and after this line. If it happens, it means that the voice recording is not original anymore. The editing could be done in the purpose of to remove unwanted parts or to add some parts. In this case, the recording could be rejected and could not be accepted to be analysed forensically because the content has been changed.

Cedar also provides spectogram for each words said. It is useful for the pruposes of voice identification or verification. With this feature, we can apply phonetics forensic in order to compare spectogram between questioned voice and known voice, so that we can know whose voice it is. In this case, we develop a technique based on FBI procedure on phonetics forensic. This procedure was described by Bruce E. Koenig on the journal "Spectographic Voice Identification: A Forensic Survey" created in 1986. In this journal he also explained that the comparison should be performed on at least 20 different words which are pronouced similarly for meaningful results. Below the complete quotation from his journal about the comparison procedure:


(1) Only original recordings of voice samples were accepted for examination, unless the original recording had been erased and a high-quality copy was still available.
(2) The recordings were played back on appropriate professional tape recorders and recorded on a professional full-track tape recorder at 7 1/2 ips. When possible, playback speed was adjusted to correct for original recording speed errors by analyzing the recorded telephone and AC line tones on spectrum analysis equipment. When necessary, special recorders were used to allow proper playback of original recordings that had incorrect track placement or
azimuth misalignment.

(3) Spectrograms were produced on Voice Identification, Inc., Sound Spectrographs, model 700. in the linear expand frequency range (0-4000 Hz), wideband filter (300 Hz) and bar display mode. All spectrograms for each separate comparison were prepared on the same spectrograph. The spectrograms were phonetically marked below each voice sound.
(4) When necessary, enhanced tape copies were also prepared from the original recordings using equalizers, notch filters, and digital adaptive predictive deconvolution programs13,14 to reduce extraneous noise and correct telephone and recording channel effects. A second set of spectrograms was then prepared from the enhanced copies and was used together with the unprocessed spectrograms for comparison.
(5) Similarly pronounced words were compared between two voice samples, with most known voice samples being verbatim with the unknown voice recording. Normally, 20 or more different words were needed for a meaningful comparison. Less than 20 words usually resulted in a less conclusive opinion, such as possibly instead of probably.
(6) The examiners made a spectral pattern comparison between the two voice samples by comparing beginning, mean and end formant frequency, formant shaping, pitch, timing, etc., of each individual word. When available, similarly pronounced words within each sample were compared to insure voice sample consistency. Words with spectral patterns that were distorted, masked ‘by extraneous sounds, too faint, or lacked adequate identifying characteristics were
not used
(7) An aural examination was made of each voice sample to determine if pattern similarities or dissimilarities noted were the product of pronunciation differences, voice disguise, obvious drug or alcohol use, altered psychological state, electronic manipulation, etc.
(8) An aural comparison was then made by repeatedly playing two voice samples simultaneously on separate tape recorders, and electronically switching back and forth between the samples while listening on high-quality headphones. When one sample had a wider frequency response than the other, bandpass filters were used to compensate during at least some of the aural listening tests.
(9) The examiner then had to resolve any differences found between the aural and spectral results, usually by repeating all or some of the comparison steps.
(10) If the examiner found the samples to be very similar (identification) or very dissimilar (elimination), an independent evaluation was always conducted by at least one, but usually two other examiners to confirm the results. If differences of opinions occurred between the examiners, they were then resolved through additional comparisons and discussions by all the examiners involved. No or low confidence decisions were usually not reviewed by another examiner.



According to his survey, only 1 false identification case (i.e. 0.31%) was found from 318 cases of phonetics forensic, while only 2 false eliminations (i.e. 0.53%) were found from 378 phonetics forensic cases. From this data, it means that the FBI technique is reliable for voice identification or verification.

In order to run this procedure of phonetics forensic, Cedar is reliable as well as noise filtering and editing line recognition.


Good luck...!



Thursday 19 November 2009

Face Sketching

This material actually is my slides presentation when being requested to be instructor on Frontline Forensic Course in Indonesia. This course has been being conducted since 16 November till 4 December 2009. In this course, I deliver teaching materials about Digital Forensic, Face Sketching, Photography Forensic, Fire Investigation and GPS. In this post, I just describe my materials on Face Sketching. The full version of this material can be downloaded at this link http://www.scribd.com/doc/22742609/Face-Sketching.

Face sketching is required when criminal investigators would like to obtain description of somebody based on the witness testimony. It is performed in order to have suspect's face so that it is easy to recognise by the investigators, even public can know him when seeing him anytime and anywhere. To reach this condition, the results of face sketching are distributed not only for law enforcement agencies but also for public. When people see the suspect, they will contact law enforcement agency to inform his existence; and then the investigators could arrest him soon.

However, there are also problems encountering the process of face sketching, such as:
1. Witnesses might saw the suspect at glance, so that it is not enough to identify him perfectly as the witness can not give face components in details.
2. Witnesses might saw the suspect from behind and or from right/left side, so that the description of suspect's face is not sufficient to describe.
3. The limits of memory of the witnesses. It means that when a witness saw the suspect, it does not guarantee he could recognise the suspect's face in details. There is possibility that the face description will be distorted as the witness can not remember each face components in details.
4. The lack of lights. When a witness saw the suspect in the condition of dark situation (i.e a little light), he can not describe the suspect's face perfectly as he can not see it well.

In the process of face sketching, there are two components which chould be considered, namely:
1. General characteristics
2. Class characteristics

In General characteristics, it shows a general pattern of a face component such as:
1. Eyes
2. Noses
3. Eyebrows
4. Hair
5. Lips
6. Head shapes
7. Jaw shapes
8. Mustaches
9. Beards
10. Eye lines
11. Smile lines

Meanwhile, in class characteristics, it shows a specific pattern of a component and or a specific sign on a face such as:
1. Mole or beauty spot
2. Cockeye or a cast in the eye
3. Harelip
4. Scar or wound print

The techniques developed in the process of face sketching are:
1. Manual. It requires a good painter

2. Automatic. It requires a reliable computer application

To obtain the description of face sketching in details, please click the link above. Hopefully it could be useful for anybody who would like to explore face recognition. Good luck...!
 

Monday 2 November 2009

Digital Forensic: State of the art

I think it is a long time for me not to post a new topic in this blog. For this reason, I apologise because I have been so busy with some crime scene processing and digital forensic analysis.

In this post, I would like to describe a more detail about digital forensic from investigation flowchart and digital forensic procedure to study case. It is in the form of a presentation which will be delivered at the British Council, Jakarta on 7 November 2009. At that moment which is 25th anniversary of British Chevening Scholarship Scheme, I am invited to deliver this topic as I was awarded Chevening scholarship when joining MSc in Forensic Informatics at the University of Strathclyde, UK in 2008/2009. This presentation can be downloaded at http://www.scribd.com/doc/22000028/Digital-Forensic-State-of-the-Art-BC071109.

On slide 3, I explain that in the investigation of the case of computer crime and computer-related crime, digital forensic gives fully technical support to criminal investigators in order to solve the case. Digital evidence found by digital forensic analyst will be basis for the investigator to decide further investigative steps. When the case is brought to the court, the forensic analyst will be requested to give expert testimony regarding the digital evidence found. If they can explain it properly, so it can be accepted by court as a strong evidence, no doubt at all.

On slide 4, it is described that digital forensic acts not only at computer crime, but also at computer-related crime. It means that digital forensic covers a wide area of investigation where computer is used. In this crime, computer has three roles, namely computer as the tool to commit the crime, computer as the target of the crime and computer as a media for storing data related to the crime.

On slide 6, the definition of digital forensic is given. It is the application of computer science and IT technology in order to solve a crime for justice purposes. Based on this definition, digital forensic plays some key roles, namely:
  1. To support and perform scientific crime investigation
  2. To perform forensic analysis on digital evidence
  3. To be able to describe a crime connection between suspect and evidence
  4. To deliver expert testimony at court.
On slide 8, I emphasize digital forensic princples which must be applied since digital forensic is performed. These principles derived from the guideline of ACPO (Association of Chief Police Officers), UK are as follows:
  1. Principle 1: No action taken by law enforcement agencies should change data held on a computer or storage media.
  2. Princple 2: The person accessing the data must be competent to do so and be able to explain the relevance and implications of the actions taken.
  3. Principle 3: An audit trail or record of all processes applied should be created and preserved.
  4. Principle 4: The person in charge has overall responsibility to ensure that these principles are adhered to.
The principles above must be applied by digital forensic analyst when performing the investigation of computer crime or computer-related crime. Once one principle is missed, so the results of digital forensic analysis becomes weak and doubted even it is possible to be refused by court. These principles are strict to implemented during the analysis.


On other slides, please download the presentation material from the link above. I hope it can be useful in positive meaning for someone who would like to apply and develop digital forensic. Good luck....!