Optimising CSRNet with parameter-free attention mechanisms for crowd counting in public transport

Occupancy estimation and crowd counting are critical tasks in designing smart and efficient public transport vehicles. Given that public transport loading can vary from sparse to crowded, classical models for occupancy estimation must be adapted to suit this purpose. Attention mechanisms have shown remarkable capability in enhancing the representational power of deep neural networks for crowd counting in congested scenes with occlusion, complex backgrounds, and perspective distortion. However, conventional approaches, often implemented as parameterized sub-networks within convolutional layers, inevitably increase model size and computational cost, limiting deployment on resource-constrained edge devices. This paper investigates the effectiveness of state-of-the-art parameter-free attention mechanisms for crowd counting and density map estimation in highly congested scenes. We evaluate channel-wise (PFCA), spatial-wise (SA), and 3-D (SimAM) modules and compare their performance with parameterized attention modules constrained to introduce no more than 1% additional parameters. Furthermore, we present a novel combination of attention mechanisms that combines the strengths of PFCA and SA (PFCASA) customized for analyzing video streams onboard public transport systems. Using CSRNet as the backbone, experiments on the ShanghaiTech dataset demonstrate that parameter-free attention mechanisms achieve comparable or superior accuracy without introducing additional model parameters. A detailed performance analysis further reveals that PFCASA outperforms other attention modules in scenes with fewer than 40 individuals, while PFCA shows greater effectiveness as crowd density increases, underscoring their potential applicability for integration into smart public transport modalities.

Artificial Plants - Vascular Morphogenesis Controller-guided growth of braided structures

Natural plants are exemplars of adaptation through self-organisation and collective decision making. As such, they provide a rich source of inspiration for adaptive mechanisms in artificial systems. Plant growth - a structure development mechanism of continuous material accumulation that expresses encoded morphological features through environmental interactions - has been extensively explored in-silico. However, ex-silico scalable morphological adaptation through material accumulation remains an open challenge. In this paper, we present a novel type of biologically inspired modularity, and an approach to artificial growth that combines the benefits of material continuity through braiding with a distributed and decentralised plant-inspired Vascular Morphogenesis Controller (VMC). The controller runs on nodes that are capable of sensing and communicating with their neighbours. The nodes are embedded within the braided structure, which can be morphologically adapted based on collective decision making between nodes. Human agents realise the material adaptation by physically adding to the braided structure according to the suggestion of the embedded controller. This work offers a novel, tangible and accessible approach to embedding mechanisms of artificial growth and morphological adaptation within physically embodied systems, offering radically new functionalities, innovation potentials and approaches to continuous autonomous or steered design that could find application within fields contributing to the built environment, such as Architecture.