Pubblicazioni
What does CLIP know about peeling a banana?
4.1 Contributo in Atti di convegnoCorrispondenze
Keyowrd: Affordances, Affordance grounding
Abstract: rely on object-interaction examples and support a closed set of actions. These limitations hinder scalability, may introduce biases, and usually restrict models to a limited set of predefined actions. This paper proposes Affordance-CLIP, to overcome these limitations by leveraging the implicit affordance knowledge embedded within large pre-trained Vision-Language models like CLIP. We experimentally demonstrate that CLIP, although not explicitly trained for affordances detection, retains valuable, Humans show an innate capability to identify tools to support specific actions. The association between objects parts and the actions they facilitate is usually named affordance. Being able to segment objects parts depending on the tasks they afford is crucial to enable intelligent robots to use objects of daily living. Traditional supervised learning methods for affordance segmentation require costly pixel-level annotations, while weakly supervised approaches, though less demanding, still
Pre- and Self-calibration of underwater cameras for photogrammetric documentation of archaeological sites
4.1 Contributo in Atti di convegnoCorrispondenze
Abstract: Underwater photogrammetry has become one of the most affordable and adopted methods for the documentation and the 3D reconstruction of submerged archaeological assets. In digital photogrammetry, images are captured to exploit (using computer vision-based procedures) their intrinsic metric contents. To preserve the metric consistency and to obtain reliable 3D metric products, this process must be followed according to photogrammetric principles that are even more important in underwater, photogrammetry. The wide diffusion of low-cost and non-metric sensors requires that some attention be given to proper geometric calibration of the employed cameras. Via calibration, it is possible to opportunely describe geometric distortions that are observable on final images due to lens shapes and construction characteristics of the cameras and the optics used in the survey operations. This research addresses the importance of pre-calibration in underwater cameras, and for this purpose, three
Learning architecture as an impure discipline. Mental and emotional dimensions of architecture
4.1 Contributo in Atti di convegnoCorrispondenze
Abstract: into account in practicing architecture and plan-ning. Concepts such as “affordance”, "atmosphere", “embodiment” and "empa-thy" become central in our way of experiencing space: along with other con-cepts such as memory, identity, imagery, they redefine the balance between reflective and pre-reflective domains. This raises questions about how to learn the impurity of architecture, which claims the development of critical devices re-quired to understand biological, psychological and cultural aspects
New prognostic neural method by discrete wavelet transforms for electromechanical flight controls affected by progressive faults
4.1 Contributo in Atti di convegnoCorrispondenze
Abstract: networks (ANN), a basic and an enhanced feedforward neural network, performing the fault detection and identification of two critical progressive faults often affecting the EMA brushless motor (i.e. turn-to-turn short circuit of a stator coil and rotor static eccentricity); in order to identify a suitable data set able to guarantee an affordable ANN classification, the said failures precursors are properly pre-processed by a Discrete Wavelet Transform, extracting several features used as input
A single RGB image based 3D object reconstruction system
4.1 Contributo in Atti di convegnoCorrispondenze
Abstract: Easy and fast digitalization of real objects is especially useful when considering augmented reality (AR) and virtual reality (VR), as reconstructed objects allow a better interaction between the real and virtual worlds than using pre-made 3D CAD models. Thanks to the ubiquity of smartphones and to the spread of immersive VR devices, the AR and VR technologies are rapidly becoming popular. However, an affordable, robust and easy to use solution for object digitalization is still missing
Cr(III) Complexes Bearing a β-Ketoimine Ligand for Olefin Polymerization: Are There Differences between Coordinative and Covalent Bonding?
1.1 Articolo in rivistaCorrispondenze
Abstract: productivity in the polymerization of cyclic olefins, affording semicrystalline oligomers, without a significant difference between 1 and 2. To gain more insight, we investigated the reaction of the complexes with the Al-cocatalyst by IR and UV-Vis spectroscopies. The results proved that, in case of 1, the Al-activator deprotonates the ligand, bringing to the formation of an active species analogous to that of 2., ß-ketoimines are extensively applied for the synthesis of organometallic complexes intended as (pre)catalysts for a variety of chemical transformations. We were interested in the synthesis of two Cr complexes bearing a simple bidentate ß-ketoimine (L), with different ligand binding modes, as well as their application as a precatalyst in the polymerization of olefins. Complex 1 (L2CrCl3) was obtained by direct reaction of L with CrCl3(THF)3, while, for the synthesis of complex 2 (LCrCl2