Ethics in Deep Learning for ISR Course
Online 300-hour course in ethics in deep learning for ISR. 6 modules covering Additional Protocol I to the Geneva Conventions (arts. 36, 48 and 57), US Department of Defense Directive 3000.09 and PyTorch. Dedicated tutor and open enrolment.
Take the controls.
Run a real engineering trade-off from this programme. Move the sliders, watch the model react in real time and beat the challenge.
What you will be able to do
Explain the applications of deep learning in intelligence, surveillance and reconnaissance.
Assess the quality, representativeness and provenance of training data.
Interpret performance metrics and the consequences of errors.
Analyse the applicable principles of distinction and precaution.
Assess the privacy risks of automated surveillance.
Propose requirements for human oversight and explainability.
Prepare an ethical assessment of an ISR system based on deep learning.
Who it's for
- Deep learning engineer who wants to specialise in Ethics in Deep Learning for ISR.
- Imagery and remote sensing analyst who wants to specialise in Ethics in Deep Learning for ISR.
- Legal adviser in international humanitarian law who wants to specialise in Ethics in Deep Learning for ISR.
- Algorithm auditor who wants to specialise in Ethics in Deep Learning for ISR.
- Researcher in responsible AI who wants to specialise in Ethics in Deep Learning for ISR.
- Verification and validation engineer who wants to specialise in Ethics in Deep Learning for ISR.
- Graduates in engineering or science seeking to enter the defence sector.
What you'll work with
- Additional Protocol I to the Geneva Conventions (arts. 36, 48 and 57)
- US Department of Defense Directive 3000.09
- NATO Principles of Responsible Use of AI (2021)
- ISO/IEC TR 24029-1 (robustness of neural networks)
- ISO/IEC TR 24027 (bias in AI systems)
- European Convention on Human Rights, art. 8
- Directive (EU) 2016/680
6 modules · 300 h
1.1Intelligence, surveillance and reconnaissance: the intelligence cycle1.2Sources: IMINT, SIGINT, SAR imagery and aerial video1.3Object detection: YOLO and Faster R-CNN1.4Segmentation and change detection in satellite imagery1.5Multisensor fusion1.6Project Maven and its controversy1.7Google employee protest in 20181.8Values at stake: security, privacy and life1.9From analysis to target selection1.10Case study: analysis of Project Maven
2.1Data labelling and annotator workers2.2Datasets: xView, DOTA and SpaceNet2.3Geographic and class bias2.4Synthetic data and its validity2.5Personal data in high-resolution imagery2.6Open-source and commercial sources2.7Security of classified data2.8Data poisoning2.9Documentation with datasheets2.10Laboratory: bias analysis of an aerial dataset
3.1Metrics: precision, recall and mAP3.2Asymmetric cost of false positives3.3Civilian-combatant confusion in imagery3.4Robustness against camouflage and deception3.5Physical adversarial attacks3.6Uncertainty and calibration3.7Domain shift between theatres3.8Operational versus laboratory evaluation3.9Red team testing3.10Practice: evaluation of a detector with a confusion matrix
4.1Principle of distinction and the duty to verify4.2Precautions in attack: Article 57 of Protocol I4.3Machine speed and human deliberation4.4Automation bias in analysts4.5Target recommendation systems: current debate4.6Responsibility of the analyst and the commander4.7Legal review under Article 364.8Decision records and traceability4.9Investigation of incidents with civilian victims4.10Case study: misidentification in a drone strike
5.1ISR on national territory and its limits5.2Wide-area persistent surveillance5.3Facial recognition and remote biometrics5.4Pattern-of-life tracking5.5Proportionality and necessity5.6Judicial and parliamentary oversight5.7Commercialisation of satellite imagery5.8Risks for journalists and NGOs5.9Use of ISR systems at borders5.10Workshop: proportionality assessment of an ISR mission
6.1System description and its mission6.2Dataset analysis6.3Performance and error assessment6.4Applicable legal analysis6.5Privacy assessment6.6Human oversight design6.7Recommendations for use6.8Audit plan6.9Report and presentation6.10Final review with the tutor
How you'll learn
Fully 100% online training on the SEIUM campus, with downloadable material, recorded classes and a dedicated technical tutor. No fixed timetable: students progress at their own pace within the intake, which makes it possible to combine study with work.
Each module combines theory, guided reading of the original standard and applied exercises. The working software is PyTorch, TensorFlow, CVAT, SHAP, in free, academic or evaluation versions. Tests are studied using real procedures for evaluating the performance and errors of vision models, audits of training datasets and analysis of surveillance cases. Assessment is continuous per module plus the final integrating project.
Build it end to end
Your final project follows the real engineering workflow, step by step, reviewed by your tutor.
Explain the applications of deep learning in intelligence, surveillance and reconnaissance.
Assess the quality, representativeness and provenance of training data.
Interpret performance metrics and the consequences of errors.
Analyse the applicable principles of distinction and precaution.
Assess the privacy risks of automated surveillance.
Where this takes you
Requirements
- No prior qualification required.
- Internet connection and a computer capable of running the programme's software.
- Sufficient Spanish to follow technical regulatory reading.
Ready to start?
Send your application today. Our admissions team replies within 72 hours with your study plan and funding options.
Frequently asked questions
The programme totals 300 hours spread over 3 to 5 months. That is between 10 and 15 hours per week on average, compatible with a full-time job.
The course works with PyTorch, TensorFlow, CVAT, SHAP. Free, academic or evaluation versions are used, and the tutor indicates the free alternative where one exists.
An in-house diploma from SEIUM Institute upon passing the continuous assessment and the final integrating project. It is in-house training, not an official qualification from the Ministry.
Yes. Modules passed in a course or postgraduate diploma in the same area are recognised in the advanced programme. Check the recognition policy when you enrol.
Related programmes
Course in Measuring Claims Response Times
Online 300-hour course in measuring claims response times. 6 modules covering Ley 50/1980 de Contrato de Seguro (arts. 18, 20 and 38), ISO 13053-1 (Six Sigma,…
Course in Environmental Regulations in Temporary Bases
Online course of 300 h in environmental regulations in temporary bases. 6 modules covering ISO 14001 (environmental management), STANAG 7141 (environmental protection in NATO military activities)…
Course in Wave and Tidal Energy: Technical Fundamentals
Online 300 h course in wave and tidal energy: technical fundamentals. 6 modules covering IEC TS 62600-101 Wave energy resource assessment and characterisation, IEC TS 62600-201…
More in Regulation, Safety, RAMS & Homologation
All Regulation, Safety, RAMS & Homologation programmes → · How to apply · Fees & financing · Scholarships