Forensic Audio, Video and Image Analysis
Authenticating images, video and audio: forgery detection, source-device identification, and recovering intelligence from degraded recordings.
- 60hours
- 32topics
- 10modules
Foundations: Digital Media and the Forensic Frame
The physical and mathematical basis of digital images, video, and audio, and the principles that allow these media to be analysed as evidence. Establishes what a digital media file actually is before authentication questions are asked.
Start module- Digital Image Fundamentals: Pixels, Sensors, and FormatsEvery digital image is a grid of numbers shaped by its sensor, demosaicing, and file format, each leaving forensically exploitable signatures.13 min
- Video and Audio File Structures: Containers, Codecs, and CompressionVideo and audio files layer a container over a codec, and the interaction between the two creates artefacts used to detect re-encoding and source devices.13 min
- Metadata, EXIF, and File-Structure ForensicsEXIF, IPTC, XMP, and structures like MP4 atoms and WAV chunks encode timestamps and device IDs that can corroborate or contradict a file's claimed origin.15 min
Image Authentication and Forgery Detection
Methods for determining whether a digital image is an authentic, unaltered record or has been manipulated. Covers the main classes of forgery and the signal-level techniques used to detect them.
Start module- Image Authentication: Principles and the Forgery TaxonomyImage authentication decides whether a photograph has been manipulated, covering the four manipulation classes and the passive versus active divide.12 min
- JPEG Compression Artefacts and Double-Compression DetectionJPEG's compression leaves a statistical fingerprint that manipulation disturbs, detectable via DCT structure, quantisation tables, and the JPEG ghost method.12 min
- Error Level Analysis (ELA)Error level analysis amplifies the per-pixel difference between an image and a freshly re-saved copy to reveal regions with inconsistent compression history.12 min
- Copy-Move and Splicing DetectionCopy-move and splicing are the most common structural forgeries in photographs, detected by block-matching, keypoint algorithms, and deep-learning detectors.12 min
- Noise and CFA Inconsistency DetectionAnalysing a camera's noise signature and CFA residuals across image regions can localise forgeries and distinguish steganography from compositing.13 min
Source Camera Identification: PRNU
Photo-Response Non-Uniformity noise as a sensor fingerprint. Covers the physics of PRNU, methods for extracting and comparing camera patterns, and the statistical framework for reaching evidential conclusions.
Start module- PRNU: Physics and the Camera FingerprintPhoto Response Non-Uniformity arises from sensor manufacturing variation, giving every camera a unique pixel pattern that can link an image to the device.13 min
- PRNU Extraction, Matching, and the PCE StatisticPRNU camera attribution denoises an image to isolate the sensor residual, then compares it to a reference pattern with the Peak-to-Correlation Energy statistic.13 min
- PRNU in Casework: Limitations and ReportingPRNU camera attribution is limited by compression and re-encoding in practice, so rigorous reporting must address false-positive risk and uncertainty.15 min
Deepfake and AI-Generated Media Detection
Detection of synthetic media produced by generative AI: face swaps, full face synthesis, voice cloning, and generative image models. Covers the generation mechanisms, current detection methods, and the limitations of the field.
Start module- Deepfake Generation: GANs, Diffusion, and Face-Swap PipelinesA technical walkthrough of how deepfakes are made, covering GAN-based face swapping, diffusion synthesis, neural talking heads, and voice cloning.13 min
- Deepfake Detection: Signal, Physiological, and Semantic MethodsA survey of how deepfake videos are detected, from frequency-domain analysis and CNN residual detectors to physiological signals like eye-blink rate.13 min
- AI-Generated Image Detection and ProvenanceHow forensic scientists detect whether an image was captured or AI-generated, covering GAN fingerprints, diffusion spectral signatures, and C2PA manifests.14 min
- Synthetic Media in Casework and Legal ProceedingsHow courts handle the authentication burden when a party alleges deepfake, covering chain of custody, expert testimony, and jurisdictional standards.13 min
Image Enhancement and Photogrammetry
Scientific limits of image enhancement, and how geometric information in images can be extracted to produce measurements of real-world dimensions, height, distance, and position.
Start module- Image Enhancement: Principles and the Resolution BarrierWhat image enhancement can and cannot do, grounded in the physics of spatial resolution, optical blur, and the hard limit on information never captured.13 min
- Photogrammetry and Metric Analysis from ImagesProjective geometry, camera calibration, and reference objects let analysts extract real-world measurements from images and CCTV with quantified uncertainty.11 min
- Height Estimation from CCTV FootageThe methods and mandatory uncertainty reporting for estimating a person's stature from CCTV, comparing reference-object and vanishing-point approaches.12 min
CCTV, DVR, and Video Analysis
Acquisition, authentication, and analysis of video evidence from CCTV systems and digital video recorders. Covers the technical characteristics of surveillance systems that affect evidential quality and the methods used to analyse footage.
Start module- CCTV and DVR Systems: Architecture and Evidence AcquisitionCCTV and DVR systems record differently whether analogue, IP-based, or hybrid, and each architecture demands a tailored acquisition strategy.13 min
- Video Authentication and Integrity VerificationVideo authentication uses hash verification, metadata cross-referencing, and frame-level discontinuity detection to expose tampering that looks intact.11 min
- Video Frame Analysis and EnhancementVideo frame analysis uses de-interlacing, multi-frame averaging, super-resolution, and colour calibration to improve clarity while keeping every step auditable.14 min
Facial Image Comparison
Methods for comparing faces in CCTV and other images with known reference photographs. Covers morphological analysis, the role and limits of automated recognition, and the standards governing expert facial comparison evidence.
Start module- Facial Image Comparison: Morphological and Holistic MethodsFacial image comparison examines still or video images to determine whether two faces belong to the same person, using landmark and holistic analysis.15 min
- Automated Facial Recognition: Role, Accuracy, and LimitationsAutomated facial recognition uses deep neural networks to score facial similarity, and knowing where these systems fail matters to any practitioner or court.17 min
Audio Forensics: Authentication and Enhancement
Methods for authenticating audio recordings and enhancing speech intelligibility, including the Electric Network Frequency method as a powerful time-stamping tool.
Start module- Audio Authentication and the Electric Network Frequency MethodThe Electric Network Frequency method turns the mains hum on any plugged-in recording into an involuntary timestamp for dating it and detecting tampering.13 min
- Audio Enhancement and Speech IntelligibilityForensic audio enhancement recovers intelligible speech from degraded recordings using noise reduction, adaptive filtering, and bandwidth extension.12 min
- Gunshot and Event Audio AnalysisForensic gunshot audio analysis reads the muzzle blast, shockwave, and impact signatures of a discharge to establish shot count, sequence, weapon, and location.11 min
- Speaker Comparison: Methods and the Expert's RoleForensic speaker comparison evaluates whether a known and an unknown voice share a common source, using phonetic, acoustic, and automatic recognition methods.12 min
Steganography and Hidden-Data Detection
Detection and extraction of information concealed within digital media files. Covers spatial and frequency-domain steganography methods, steganalysis techniques, and the investigative context in which hidden data appears.
Start module- Steganography Methods in Images, Audio, and VideoA practical guide to how hidden data is embedded inside images, audio, and video, covering the main spatial, transform-domain, and metadata techniques used.15 min
- Steganalysis: Statistical Detection MethodsHow examiners detect steganographic content using statistical tests and machine-learning models, from the chi-squared attack to deep-learning steganalysers.15 min
Standards, Chain of Custody, and Court Presentation
The procedural and legal framework governing media evidence: SWGDE and SWGIT guidelines, chain-of-custody requirements, report writing, and the admissibility of multimedia evidence in different legal systems.
Start module- SWGDE and SWGIT Standards for Digital and Imaging EvidenceSWGDE and SWGIT are the US working groups behind the consensus best-practice documents governing how digital multimedia evidence is acquired and analysed.11 min
- Chain of Custody for Media EvidenceMaintaining an unbroken, documented chain of custody for digital media evidence, from seizure through every processing step to its presentation in court.12 min
- Multimedia Evidence: Admissibility and Expert TestimonyHow multimedia evidence is authenticated and admitted across major legal systems, the presenting expert's obligations, and the cases that set current standards.14 min