arXiv subjects

Simone Rossetti

Publications and source records attributed to Simone Rossetti.

3 recordsLinked to original sources

arXiv2023

Removing supervision in semantic segmentation with local-global matching and area balancing

Removing supervision in semantic segmentation is still tricky. Current approaches can deal with common categorical patterns yet resort to multi-stage architectures. We design a novel end-to-end model leveraging local-global patch matching to predict categories, good localization, area and shape of objects for semantic segmentation. The local-global matching is, in turn, compelled by optimal transport plans fulfilling area constraints nearing a solution for exact shape prediction. Our model attains state-of-the-art in Weakly Supervised Semantic Segmentation, only image-level labels, with 75% mIoU on PascalVOC2012 val set and 46% on MS-COCO2014 val set. Dropping the image-level labels and clustering self-supervised learned features to yield pseudo-multi-level labels, we obtain an unsupervised model for semantic segmentation. We also attain state-of-the-art on Unsupervised Semantic Segmentation with 43.6% mIoU on PascalVOC2012 val set and 19.4% on MS-COCO2014 val set. The code is available at https://github.com/deepplants/PC2M.

cs.CV↗

arXiv2022

Max Pooling with Vision Transformers reconciles class and shape in weakly supervised semantic segmentation

Weakly Supervised Semantic Segmentation (WSSS) research has explored many directions to improve the typical pipeline CNN plus class activation maps (CAM) plus refinements, given the image-class label as the only supervision. Though the gap with the fully supervised methods is reduced, further abating the spread seems unlikely within this framework. On the other hand, WSSS methods based on Vision Transformers (ViT) have not yet explored valid alternatives to CAM. ViT features have been shown to retain a scene layout, and object boundaries in self-supervised learning. To confirm these findings, we prove that the advantages of transformers in self-supervised methods are further strengthened by Global Max Pooling (GMP), which can leverage patch features to negotiate pixel-label probability with class probability. This work proposes a new WSSS method dubbed ViT-PCM (ViT Patch-Class Mapping), not based on CAM. The end-to-end presented network learns with a single optimization process, refined shape and proper localization for segmentation masks. Our model outperforms the state-of-the-art on baseline pseudo-masks (BPM), where we achieve $69.3\%$ mIoU on PascalVOC 2012 $val$ set. We show that our approach has the least set of parameters, though obtaining higher accuracy than all other approaches. In a sentence, quantitative and qualitative results of our method reveal that ViT-PCM is an excellent alternative to CNN-CAM based architectures.

cs.CV↗

arXiv2020

Pencil Beamforming Increases Human Exposure to ElectroMagnetic Fields: True or False?

According to a very popular belief - very widespread among non-scientific communities - the exploitation of narrow beams, a.k.a. "pencil beamforming", results in a prompt increase of exposure levels radiated by 5G Base Stations (BSs). To face such concern with a scientific approach, in this work we propose a novel localization-enhanced pencil beamforming technique, in which the traffic beams are tuned in accordance with the uncertainty localization levels of User Equipment (UE). Compared to currently deployed beamforming techniques, which generally employ beams of fixed width, we exploit the localization functionality made available by the 5G architecture to synthesize the direction and the width of each pencil beam towards each served UE. We then evaluate the effectiveness of pencil beamforming in terms of ElectroMagnetic Field (EMF) exposure and UE throughput levels over different realistic case-studies. Results, obtained from a publicly released open-source simulator, dispel the myth: the adoption of localization-enhanced pencil beamforming triggers a prompt reduction of exposure w.r.t. other alternative techniques, which include e.g., beams of fixed width and cellular coverage not exploiting beamforming. The EMF reduction is achieved not only for the UE that are served by the pencil beams, but also over the whole territory (including the locations in proximity to the 5G BS). In addition, large throughput levels - adequate for most of 5G services - can be guaranteed when each UE is individually served by one dedicated beam.

cs.NI↗