Scientists at MIT's LabSix,erotice يابانية اÙلام an artificial intelligence research group, tricked Google's image-recognition AI called InceptionV3 into thinking that a baseball was an espresso, a 3D-printed turtle was a firearm, and a cat was guacamole.
The experiment might seem outlandish initially, but the results demonstrate why relying on machines to identify objects in the real world could be problematic. For example, the cameras on self-driving cars use similar technology to identify pedestrians while in motion and in all sorts of weather conditions. If an image of a stop sign was blurred (or altered), an AI program controlling a vehicle could theoretically misidentify it, leading to terrible outcomes.
The results of the study, which were published online today, show that AI programs are susceptible to misidentifying objects in the real-world that are slightly distorted, whether manipulated intentionally or not.
SEE ALSO: After getting shade from a robot, Elon Musk fires backAI scientists call these manipulated objects or images, such as turtle with a textured surface that might mimic the surface of a rifle, "adversarial examples."
"Our work demonstrates that adversarial examples are a significantly larger problem in real world systems than previously thought," the scientists wrote in the published research.
The example of the 3D-printed turtle below proves their point. In the first experiment, the team presents a typical turtle to Google's AI program, and it correctly classifies it as a turtle. Then, the researchers modify the texture on the shell in minute ways — almost imperceptible to the human eye — which makes the machine identify the turtle as a rifle.
The striking observation in LabSix's study is that the manipulated or "perturbed" turtle was misclassified at most angles, even when they flipped the turtle over.
To create this nuanced design trickery, the MIT researchers used their own program specifically designed to create "adversarial" images. This program simulated real-world situations like blurred or rotating objects that an AI program could likely experience in the real-world — perhaps like the input an AI might get from cameras on fast-moving self-driving cars.
With the seemingly incessant progression of AI technologies and their application in our lives (cars, image generation, self-taught programs), it's important that some researchers are attempting to fool our advanced AI programs; doing so exposes their weaknesses.
After all, you wouldn't want a camera on your autonomous vehicle to mistake a stop sign for a person — or a cat for guacamole.
Topics Artificial Intelligence Google
Zoom announces AI features that act as your personal assistantAs New York City schools ban Zoom, online privacy resonates with video'The Mandalorian' brings Zeb Orrelios from 'Star Wars Rebels' into live actionRise of AI: The internet isn't owned by users anymoreNeed to switch up your Zoom background? We made you 15 delightful new options.Levi's to use AI models to 'increase diversity'Now might be a good time to finally try a retinoidWhy we enjoy being humiliated during sexWordle today: Here's the answer, hints for March 29'Succession' Season 4, episode 1: Logan's phone call scene, explained'Bheed' actors speak about COVIDDisease experts weigh in on crowded White House coronavirus briefingsDog and cat Google searches surge to new high during COVIDThe best tweets of this week, because why not?Down Dog app review: At'Yellowjackets' connects Seasons 1 and 2 with 'Beaches'10 things to fantasize about doing once coronavirus is over'Yellowjackets' connects Seasons 1 and 2 with 'Beaches'Social worker urges safe distancing with face mask covered in penisesFightCamp review: We're all anxious right now and this at Japan’s Nissan receives more than 20,000 non NIO Onvo SUVs reportedly bought for high Four Chinese companies plan to build EV Li Auto reportedly cuts delivery goal amid fierce competition, lackluster demand · TechNode CATL to offer skateboard chassis to Mazda's China joint venture · TechNode China’s Xpeng showcases EVs at Milan Design Week, makes foray into Italy · TechNode JD.com’s food delivery service faces backlash after system crash amid 618 surge · TechNode US ends duty Foreign tourist spending on Alipay surges in China over May Day Holiday · TechNode Tencent not in talks to acquire Nexon, source says: report · TechNode TSMC nears 2nm rollout with per wafer prices climbing to $30,000 · TechNode Alibaba CEO urges re Oppo signs global patent deal with Volkswagen, expanding 5G tech to connected cars · TechNode BYD surpasses Tesla in Europe EV sales for the first time: JATO · TechNode TSMC to open Munich Design Centre in Q3 to support advanced chip design for Europe · TechNode DeepSeek reveals cost China’s GAC starts operations at new Indonesia EV plant · TechNode Qualcomm bets on on NetEase to launch mobile adaptation of survival game Frostpunk tomorrow · TechNode Shanghai cracks down on illegal AI content on major platforms · TechNode
3.532s , 8286.6640625 kb
Copyright © 2025 Powered by 【erotice يابانية اÙلام】,Feast Information Network