Os softwares de reconhecimento facial estão cada vez mais avançados e precisos, mas uma análise realizada por pesquisadores da Universidade Stanford e do Instituto de Tecnologia de Massachussetts (MIT) com três programas disponíveis no mercado mostra que a raça e o sexo influenciam os resultados. Em testes com homens brancos, o erro nunca foi superior a 0,8%, mas com mulheres negras, as taxas de erro foram de 20% num software e de mais de 34% nos outros dois.Os resultados levantam questões sobre como as redes neurais são treinadas pelas empresas que desenvolvem os softwares. Por meio da inteligência artificial, esses programas “aprendem” ao analisar toneladas de dados. O problema é que o material apresentado para a máquina aprender está enviesado. Uma das companhias avaliadas alega que o sistema tem precisão acima de 97%, mas a base de dados utilizada para alimentá-lo era composta 77% por homens e 83%, brancos. E essa falha pode ter repercussões graves.Disponível em: <https://epocanegocios.globo.com/Tecnologia/noticia/2018/02/softwares-de-reconhecimento-facial-funcionam-bem-mas-apenas-para-homens-brancos.html>. Acesso em: jan. 2024. (Adaptado.)O texto levanta questões sobre o treinamento das redes neurais por empresas desenvolvedoras de softwares de reconhecimento facial. Pela leitura, infere-se quea)os resultados imprecisos dos softwares de reconhecimento facial para mulheres negras e outros grupos minoritários são resultado de falhas técnicas nos algoritmos utilizados, não do treinamento das redes neurais.b)o treinamento das redes neurais dos softwares de reconhecimento facial não influencia os resultados obtidos, que são determinados apenas pela qualidade dos algoritmos, independentemente de raça ou gênero.c)as redes neurais dos softwares de reconhecimento facial são treinadas com base em uma diversidade de dados que reflete a diversidade populacional, o que garante resultados precisos para todas as raças e gêneros.d)a falta de diversidade nos dados utilizados para treinar as redes neurais dos softwares de reconhecimento facial pode gerar resultados enviesados e menos precisos para determinados grupos raciais e de gênero.e)os softwares de reconhecimento facial apresentam precisão igual para todas as raças e gêneros, independentemente dos dados utilizados em seu treinamento, haja vista a diversidade de dados empregados em sua base.

Quelle technologie ML permet d'identifier et de suivre des individus spécifiques ?Veuillez choisir une réponse.détection de la fraude à la carte bancaireReconnaissance facialelecteurs en ligneReconnaissance de forme

Do you have experience with Facial landmark extraction and head pose estimation algorithms/modules?  If yes, please mention the project in brief.

Face recognition system is based on _____________.

A startup company called Qoves has developed an AI-powered beauty assessment tool that tells you how attractive you are. The free version spits out a list of your “predicted flaws'' and explains what sort of surgical interventions and expensive serums are needed to “fix” you. They’re not the only ones playing around with AI for this purpose, there are an alarming number of these sort of beauty assessment algorithms online. Face++, a facial recognition platform developed by China’s Megvii Technology, for example, has a tool that gives you a “Beauty Score”. It also purports to tell you how attractive you are from both a male and female perspective.How does a beauty scoring algorithm work? Good question; nobody really knows. Face++ won’t reveal the workings of its algorithm. Qoves’s tool, meanwhile, is trained on a dataset of hundreds of thousands of pictures scored manually by humans and extrapolates attractiveness from that. One thing we can say for certain about these sorts of tools is that they often reflect Eurocentric beauty biases. 2016, for example, saw the world’s first international beauty contest judged entirely by algorithm. Guess what? Out of 44 winners, nearly all were white, a handful were Asian, and only one had dark skin.In no more than two sentences, discuss why using an AI tool to rate attractiveness, for a sample of Unsplash photos of people, may be problematic. In your response include at least two relevant things you learned from this article.

Analysing Facial ExpressionsSuppose you have sent an email to one of your subordinates asking him to send you a report urgently. You give him until the end of the day to send the report, but do not receive a response from him even after the day passes. The next day, you call him to your cabin and ask him for the report. In a surprised tone, he asks you, “Which report?”.He raises his eyebrows and they get drawn together. Horizontal wrinkles appear on his forehead and only the upper white portion of his eyeballs becomes visible.You check your mailbox to see if you had somehow missed sending him the email. The email is there in your ‘Sent’ box, and so you know that the communication was proper from your end.How should you respond in this situation? (Answer the question based on the concepts you learnt in the videos above.)Give him one more day to submit the report, as you can tell from his expression that he was genuinely surprised by the mention of the report and must have missed your email.✕ IncorrectFeedback:In the microexpression of surprise, both the upper and lower white portions of the eyeballs become visible. However, in this situation, only the upper white portion of your subordinate’s eyeballs is visible. Hence, this option is incorrect.Reprimand him for the missed deadline, because you know he is not telling the truth and has indeed received your email.Assign the task to someone else, as you can tell from his expression that he is extremely saddened by some event.Let him calm down first and then ask for the report because he seems quite angry.