Hyperdimensional computing (HDC) offers lightweight learning for energy-constrained devices by encoding data into high-dimensional vectors.However, its reliance on ultra-high dimensionality and static, randomly initialized hypervectors limits memory efficiency and learning capacity.Therefore, we propose Trainable Hyperdimensional Computing (THDC), which enables end-to-end HDC via backpropagation.THDC replaces randomly initialized vectors with trainable embeddings and introduces a one-layer binary neural network to optimize class representations.Evaluated on MNIST, Fashion-MNIST and CIFAR-10, THDC achieves equal or better accuracy than state-of-the-art HDC, with dimensionality reduced from 10.000 to 64.

Multi-camera computer vision in industry offers advantages but poses risks to worker privacy and intellectual property through exposure of sensitive contextual information. Existing privacy methods often inadequately protect background details crucial in manufacturing. This issue is prominent in applications like automated ergonomic assessment, where visual data for posture analysis can reveal sensitive workplace information. We propose a system for simultaneous personal privacy and enhanced contextual intellectual property protection, featuring a novel probabilistic obfuscation technique. Our edge-based Generative Adversarial Privacy system employs a modified obfuscator that learns to inject controlled, pixel-wise random noise, particularly into non-critical background regions. This more effectively obscures IP-sensitive environmental details before data transmission for central analysis (e.g., pose estimation). Our approach, validated in a multi-camera ergonomic study, effectively protects worker privacy and contextual IP (metrics-evaluated) and maintains 3D pose accuracy for reliable ergonomic assessment. This work provides a solution for deploying vision systems in sensitive industrial settings by holistically addressing privacy requirements through an advanced, adaptive obfuscation strategy.

Specifying tasks for robotic systems traditionally requires coding expertise, deep domain knowledge, and significant time investment. While learning from demonstration offers a promising alternative, existing methods often struggle with tasks of longer horizons. To address this limitation, we introduce a computationally efficient approach for learning probabilistic deterministic finite automata (PDFA) that capture task structures and expert preferences directly from demonstrations. Our approach infers sub-goals and their temporal dependencies, producing an interpretable task specification that domain experts can easily understand and adjust. We validate our method through experiments involving object manipulation tasks, showcasing how our method enables a robot arm to effectively replicate diverse expert strategies while adapting to changing conditions.

This paper presents an end-to-end, IoT-enabled robotic system for the non-destructive, real-time, and spatially-resolved mapping of grape yield and quality (Brix, Acidity) in vineyards. The system features a comprehensive analytical pipeline that integrates two key modules: a high-performance model for grape bunch detection and weight estimation, and a novel deep learning framework for quality assessment from hyperspectral (HSI) data. A critical barrier to in-field HSI is the “domain shift" caused by variable illumination. To overcome this, our quality assessment is powered by the Light-Invariant Spectral Autoencoder (LISA), a domain-adversarial framework that learns illumination-invariant features from uncalibrated data. We validated the system’s robustness on a purpose-built HSI dataset spanning three distinct illumination domains: controlled artificial lighting (lab), and variable natural sunlight captured in the morning and afternoon. Results show the complete pipeline achieves a recall (0.82) for bunch detection and aR2(0.76) for weight prediction, while the LISA module improves quality prediction generalization by over 20% compared to the baselines. By combining these robust modules, the system successfully generates high-resolution, georeferenced data of both grape yield and quality, providing actionable, data-driven insights for precision viticulture.

Abstract Visual analysis tasks, including crowd management, often require resource-intensive machine learning models, posing challenges for deployment on edge hardware. Consequently, cloud computing emerges as a prevalent solution. To address privacy concerns associated with offloading video data to remote cloud platforms, we present a novel approach using adversarial training to develop a lightweight obfuscator neural network. Our method focuses on pedestrian detection as an example of visual analysis, allowing the transformation of video frames on the camera itself to retain only essential information for pedestrian detection while preserving privacy. Importantly, the obfuscated data remains compatible with publicly available object detectors, requiring no modifications or significant loss in accuracy. Additionally, our technique overcomes the common limitation of relying on labeled sensitive attributes for privacy preservation. By demonstrating the inability of pedestrian attribute recognition models to detect attributes in obfuscated videos, we validate the efficacy of our privacy protection method. Our results suggest that this scalable approach holds promise for enabling camera usage in video analytics while upholding personal privacy.

Automated anomaly detection in surveillance videos has attracted much interest as it provides a scalable alternative to manual monitoring. Most existing approaches achieve good performance on clean benchmark datasets recorded in well-controlled environments. However, detecting anomalies is much more challenging in the real world. Adverse weather conditions like rain or changing brightness levels cause a significant shift in the input data distribution, which in turn can lead to the detector model incorrectly reporting high anomaly scores. Additionally, surveillance cameras are usually deployed in evolving environments such as a city street of which the appearance changes over time because of seasonal changes or roadworks. The anomaly detection model will need to be updated periodically to deal with these issues. In this paper, we introduce a multi-branch model that is equipped with a trainable preprocessing step and multiple identical branches for detecting anomalies during day and night as well as in sunny and rainy conditions. We experimentally validate our approach on a distorted version of the Avenue dataset and provide qualitative results on real-world surveillance camera data. Experimental results show that our method outperforms the existing methods in terms of detection accuracy while being faster and more robust on scenes with varying visibility.

As warehouses, storage facilities and factories become more expanded and equipped with smart devices, there is a substantial need for rapid, intelligent and autonomous detection of unusual and potentially hazardous situations, also called anomalies. In particular for Autonomous Guided Vehicles (AGVs) that drive around these premises independently, unforeseen obstructions along their path-e.g. a cardboard box in the middle of a corridor or bumps in the floor-and sudden or unexpected actions executed by personnel-e.g. someone walking in a restricted area-make it hard for AGVs to navigate safely. We therefore propose a novel approach to detect such anomalies in an unsupervised manner by measuring Bayesian surprise: whenever an event is observed that does not align with the agent's prior knowledge of the world, this event is deemed surprising and could indicate an anomaly. This paper lays out the details on how to learn both the prior and posterior models of an AGV that drives around a warehouse and observes the environment through an RGBD camera. In the experiments we show that our Bayesian surprise approach outperforms a baseline that is traditionally used to detect anomalies in sequences of images.

Internet applications rely on strong encryption techniques to protect the content of all communications between client and server. These encryption algorithms ensure that third parties are unable to obtain the plain text data but also make it hard for the network administrator to enforce restrictions on the types of traffic that are allowed. In this paper we show that we can train accurate machine learning models which can predict the type of traffic going through an IPsec or TOR tunnel based on features extracted from the encrypted streams. We use small, fast to execute machine learning models that work on small windows of data. This makes it possible to use our approach in real-time, for example as part of a Quality of Service (QoS) system.

Using deep neural networks on resource constrained devices is a trending topic in neural network research. Various techniques for compressing neural networks have been proposed that allow evaluating a large neural network on a device with limited memory and processing power. These approaches usually generate a single compressed student network based on a larger teacher network. In some cases a more dynamic trade-off may be desired. In this paper we trained a sequence of increasingly large networks where each network is constrained to contain the unmodified features of all smaller networks. The weight matrix of the largest network has submatrices that correspond to the weight matrices of each of the smaller networks. This technique allows us to keep the parameters of several networks in memory while having the same memory footprint as the single largest network. A trade-off between accuracy and speed can be made at runtime. The proposed approach is validated on two image classification tasks running on a real-world Internet-of-Things (IoT) device.

Deep neural networks are the state of the art technique for a wide variety of classification problems. Although deeper networks are able to make more accurate classifications, the value brought by an additional hidden layer diminishes rapidly. Even shallow networks are able to achieve relatively good results on various classification problems. Only for a small subset of the samples do the deeper layers make a significant difference. We describe an architecture in which only the samples that can not be classified with a sufficient confidence by a shallow network have to be processed by the deeper layers. Instead of training a network with one output layer at the end of the network, we train several output layers, one for each hidden layer. When an output layer is sufficiently confident in this result, we stop propagating at this layer and the deeper layers need not be evaluated. The choice of a threshold confidence value allows us to trade-off accuracy and speed.