Filters








51 Hits in 0.79 sec

Supervised Multimodal Bitransformers for Classifying Images and Text [article]

Douwe Kiela, Suvrat Bhooshan, Hamed Firooz, Ethan Perez, Davide Testuggine
2020 arXiv   pre-print
Self-supervised bidirectional transformer models such as BERT have led to dramatic improvements in a wide variety of textual classification tasks. The modern digital world is increasingly multimodal, however, and textual information is often accompanied by other modalities such as images. We introduce a supervised multimodal bitransformer model that fuses information from text and image encoders, and obtain state-of-the-art performance on various multimodal classification benchmark tasks,
more » ... forming strong baselines, including on hard test sets specifically designed to measure multimodal performance.
arXiv:1909.02950v2 fatcat:ujdciqsh5faq3jh2yd4lopraae

The Hateful Memes Challenge: Detecting Hate Speech in Multimodal Memes [article]

Douwe Kiela, Hamed Firooz, Aravind Mohan, Vedanuj Goswami, Amanpreet Singh, Pratik Ringshia, Davide Testuggine
2021 arXiv   pre-print
This work proposes a new challenge set for multimodal classification, focusing on detecting hate speech in multimodal memes. It is constructed such that unimodal models struggle and only multimodal models can succeed: difficult examples ("benign confounders") are added to the dataset to make it hard to rely on unimodal signals. The task requires subtle reasoning, yet is straightforward to evaluate as a binary classification problem. We provide baseline performance numbers for unimodal models,
more » ... well as for multimodal models with various degrees of sophistication. We find that state-of-the-art methods perform poorly compared to humans (64.73% vs. 84.7% accuracy), illustrating the difficulty of the task and highlighting the challenge that this important problem poses to the community.
arXiv:2005.04790v3 fatcat:zns5avehnrey7b2n7bpi6wbkaa

Opacus: User-Friendly Differential Privacy Library in PyTorch [article]

Ashkan Yousefpour, Igor Shilov, Alexandre Sablayrolles, Davide Testuggine, Karthik Prasad, Mani Malek, John Nguyen, Sayan Ghosh, Akash Bharadwaj, Jessica Zhao, Graham Cormode, Ilya Mironov
2021 arXiv   pre-print
We introduce Opacus, a free, open-source PyTorch library for training deep learning models with differential privacy (hosted at https://opacus.ai). Opacus is designed for simplicity, flexibility, and speed. It provides a simple and user-friendly API, and enables machine learning practitioners to make a training pipeline private by adding as little as two lines to their code. It supports a wide variety of layers, including multi-head attention, convolution, LSTM, and embedding, right out of the
more » ... ox, and it also provides the means for supporting other user-defined layers. Opacus computes batched per-sample gradients, providing better efficiency compared to the traditional "micro batch" approach. In this paper we present Opacus, detail the principles that drove its implementation and unique features, and compare its performance against other frameworks for differential privacy in ML.
arXiv:2109.12298v3 fatcat:mp4zilt5lbhjjbivwnanxkq43e

Adversarial Evaluation of Multimodal Models under Realistic Gray Box Assumption [article]

Ivan Evtimov, Russel Howes, Brian Dolhansky, Hamed Firooz, Cristian Canton Ferrer
2021 arXiv   pre-print
Online; accessed 29 October 2020. 1 [18] Douwe Kiela, Suvrat Bhooshan, Hamed Firooz, and Davide Testuggine.  ...  Generating natural language adversarial examples. arXiv preprint arXiv:1804.07998, 2018. 2 [2] Anish Athalye, Nicholas Carlini, and David Wagner.  ... 
arXiv:2011.12902v3 fatcat:y3mpjxkmoba5tng5ufjqm3z5by

Captum: A unified and generic model interpretability library for PyTorch [article]

Narine Kokhlikyan, Vivek Miglani, Miguel Martin, Edward Wang, Bilal Alsallakh, Jonathan Reynolds, Alexander Melnikov, Natalia Kliushkina, Carlos Araya, Siqi Yan, Orion Reblitz-Richardson
2020 arXiv   pre-print
Davide Testuggine for his support and initial discussions on model interpretability and integrated gradients.  ... 
arXiv:2009.07896v1 fatcat:6lnifglxpveqjbp7ij7uks4xwu

Osservatorio Erpetologico Italiano

Osservatorio Erpetologico Italiano
2019
di specie endemica italica +RAR/S Riconferma di segnalazione/sito storico (non confermato da oltre 10 anni) di specie endemica rara o al limite di areale +TES Sito accertato di riproduzione di testuggine  ...  R0010 Fiacchini David, Foglia Gessica +H-IV 110.393.0.001.0 Coronella austriaca. Civita (CS) settembre 2005. Località Colle S. Martino, 1000 m.  ... 
doi:10.13128/acta_herpetol-1883 fatcat:vdyhlk3jovck7dlpdqgjknutme

A Multimodal Framework for the Detection of Hateful Memes [article]

Phillip Lippe, Nithin Holla, Shantanu Chandra, Santhosh Rajamanickam, Georgios Antoniou, Ekaterina Shutova, Helen Yannakoudakis
2020 arXiv   pre-print
Testuggine. 2020.  ...  In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR). [14] Douwe Kiela, Suvrat Bhooshan, Hamed Firooz, and Davide Testuggine. 2019.  ... 
arXiv:2012.12871v2 fatcat:rttlifokijczthrcgduaerswey

Osservatorio Erpetologico Italiano

2019
Presso Traschio, lungo la statale 45, km 70.R0037Corpo Forestale dello Stato in servizio nel Parco Nazionale delle Cinque Terre +TES Sito accertato di riproduzione di testuggine del genere Testudo +  ...  R0047 Fiacchini David, Foglia Gessica +RAR 110.378.0.001.0 Hemidactylus turcicus. Diamante (CS), 2005. Località Cirella di Diamante.  ... 
doi:10.13128/acta_herpetol-2687 fatcat:krm5xnm74rcejixken2x5mlshy

Un'"apparizione meravigliosa, quasi Inverosimile": Tracce di musa nei versi di In un giardino 'italiano'

M Sonzogni
2011 Italian Studies in Southern Africa/Studi d'Italianistica nell'Africa Australe  
Nell'unica redazione manoscritta, tra l'altro, si trova "tartaruga" in lieu di "testuggine" (L'opera in versi:294 e 982).  ...  Ringrazio Davide Manenti per la preziosa assistenza nella compilazione di questa scheda sulla simbologia della tartaruga.  ... 
doi:10.4314/issa.v24i1.66620 fatcat:p4bcluv5pjallb52eg3gmk6n4e

Osservatorio Erpetologico Italiano

2019
Reptilia R0049 David Fiacchini +H-II +ECO 110.399.0.004.0 R0050 David Fiacchini +ECO 110.397.0.001.0 Natrix maura. Castelnovo ne' Monti (RE), 12 agosto 2007.  ...  di specie endemica italica +RAR/S Riconferma di segnalazione/sito storico (non confermato da oltre 10 anni) di specie endemica rara o al limite di areale +TES Sito accertato di riproduzione di testuggine  ... 
doi:10.13128/acta_herpetol-3424 fatcat:vasivplv3fgjdjf5mascodmmy4

MOMENTA: A Multimodal Framework for Detecting Harmful Memes and Their Targets [article]

Shraman Pramanick, Shivam Sharma, Dimitar Dimitrov, Md Shad Akhtar, Preslav Nakov, Tanmoy Chakraborty
2021 arXiv   pre-print
Douwe Kiela, Suvrat Bhooshan, Hamed Firooz, Ethan Perez, and Davide Testuggine. 2019. Supervised multimodal bitransformers for classifying images and text.  ...  Douwe Kiela, Hamed Firooz, Aravind Mohan, Vedanuj Goswami, Amanpreet Singh, Pratik Ringshia, and Davide Testuggine. 2020. The hateful memes chal- lenge: Detecting hate speech in multimodal memes.  ... 
arXiv:2109.05184v2 fatcat:ntmq4pv6kjdhvebjyohuikqppe

Hateful Memes Challenge: An Enhanced Multimodal Framework [article]

Aijing Gao, Bingjun Wang, Jiaqi Yin, Yating Tian
2021 arXiv   pre-print
Testuggine.  ...  Pre-training of deep bidirectional trans- formers for language understanding, 2019. 3 [3] Douwe Kiela, Hamed Firooz, Aravind Mohan, Vedanuj Goswami, Amanpreet Singh, Pratik Ringshia, and Davide  ... 
arXiv:2112.11244v1 fatcat:lga6pjxoe5dbzdv33mge5wlyea

SemEval-2021 Task 6: Detection of Persuasion Techniques in Texts and Images [article]

Dimitar Dimitrov, Bishr Bin Ali, Shaden Shaar, Firoj Alam, Fabrizio Silvestri, Hamed Firooz, Preslav Nakov, Giovanni Da San Martino
2021 arXiv   pre-print
Douwe Kiela, Suvrat Bhooshan, Hamed Firooz, and Davide Testuggine. 2019. Supervised multimodal bitransformers for classifying images and text.  ...  Douwe Kiela, Hamed Firooz, Aravind Mohan, Vedanuj Goswami, Amanpreet Singh, Pratik Ringshia, and Davide Testuggine. 2020. The hateful memes chal- lenge: Detecting hate speech in multimodal memes.  ... 
arXiv:2105.09284v1 fatcat:2fpwudbkebebvdfa6iry4rsunq

Corpi di scarto: la vita in discarica narrata da Elisabetta Bucciarelli // Corpi di scarto: Life in a Landfill Narrated by Elisabetta Bucciarelli // Corpi di scarto: La vida en un vertedero narrada por Elisabetta Bucciarelli

Anna Chiafele
2017 Ecozon@  
Il Vecchio viene descritto come un oggetto,"un fagotto" (20) dotato di "vita propria" (20); più spesso, egli si presenta come "una testuggine" (147).  ...  Il Vecchio, spesso paragonato ad una testuggine, produce per lo più deiezione; egli non esala "biossido di carbonio né diossina, solo una gran puzza" (175).  ... 
doi:10.37536/ecozona.2017.8.2.1428 fatcat:fypk2shf7bfzheilneya6dkokm

On the Importance of Difficulty Calibration in Membership Inference Attacks [article]

Lauren Watson and Chuan Guo and Graham Cormode and Alex Sablayrolles
2022 arXiv   pre-print
Ashkan Yousefpour, Igor Shilov, Alexandre Sablayrolles, Davide Testuggine, Karthik Prasad, Mani Malek, John Nguyen, Sayan Ghosh, Akash Bharadwaj, Jessica Zhao, et al.  ... 
arXiv:2111.08440v2 fatcat:uimf5lqkiramvohoxm7eg2hnim
« Previous Showing results 1 — 15 out of 51 results