Amazfit may not be a household name, but its been crushing it in the budget smartwatch space for a long time. And at CES 2025, the company is trying to outdo itself with the $99 Amazfit Active 2.I was a big fan of the companys GTR 4 watch it wasnt a looker, but it packed in an incredible amount of value for $200. I liked it so much that its sat comfortably as the top overall pick in our fitness tracker buying guide for two years. Compared to that watch, the Active 2 has even more functionality forhalf that price. Its got a bright 2,000 nit OLED display, upgraded health sensors, revamped heart rate and sleep algorithms, and the Zepp Flow AI assistant. Theres also a smarter strength training mode that can auto-detect sets, reps, and rest, offline maps with turn-by-turn directions, and personalized training plans via the Zepp Coach integration. For people with periods, theres also a mini app in collaboration with Wild.AI, which delivers diet and exercise recommendations based on hormonal and menstrual cycles throughout a persons life (i.e., reproductive years, perimenopause, and menopause). The watch also works in tandem with the companys recently launched Helio smart ring, similar to how the Samsung Galaxy Ring optimizes sensor accuracy with its Galaxy Watches. European users will also get contactless payments via Zepp Pay, a feature that was notably missing from the GTR 4.The display gets up to 2,000 nits of brightness.You almost never find this kind of feature set for under $250. Its bonkers that you can get it for $100. Amazfits Active line, which launched last year, effectively replaces its popular GTR lineup (though you can still easily find GTR 4 models online). Style-wise, you can see that in how the Active 2 switches from the original Actives square, Apple Watch-esque design to a more classic 44mm round watchface. The standard model comes with a tempered glass screen, stainless steel case, and a 20mm silicone sport band. The premium model costs $129 and upgrades the screen to a more durable sapphire glass and includes an additional leather strap. Compared to the Active, the Active 2 also adds more advanced hardware like a barometer, a gyroscope, and an ambient light sensor. Battery life isnt quite as good, dropping to 10 days with normal use compared to 14 days. (That said, 10 days is still impressive compared to the competition.) I wish the touchscreen were a little more responsive, but thats been my only real complaint so far.On paper, the Active 2 feels like it ought to have some kind of catch. However, Ive been wearing the premium Active 2 for the last few days, and its quite nice. So far, performance has been zippy, its lightweight on the wrist, and Ive already gotten several compliments from coworkers. Sometimes budget devices can look chintzy, and while no one would mistake the Active 2 for a luxury watch, it doesnt look or feel as cheap as a $100$150 smartwatch could. Im still testing out the health and fitness features, but so far, its held up well on a handful of treadmill runs, and in the last three days, Ive only used about 25 percent battery. I wish the touchscreen were a bit more responsive, but its nothing too egregious thus far. Its too early for me to make a definitive call on the Active 2, but so far, its looking like it could be a compelling option for folks who want a simpler smartwatch without breaking the bank.The Active 2 is available for preorder in the US now and will launch globally in February.

Sony is launching the Xyn (pronounced zin), an extended reality headset for creating 3D objects for films, animation, and games. The device, which Sony first teased at CES last year, will come with 4K OLED displays, video passthrough, and support for third-party design, modeling, and production software.Based on the video shared by Sony, it looks like wearers will be able to flip up the headset when theyre not using it. It doesnt appear to come with the ring and pointer that Sony showed off last year, though.Alongside the headset, Sony revealed Xyn Motion Studio, an accompanying Windows app that supports up to 12 Mocopi sensors, its wireless motion-tracking system that V-Tubers and other creators often use to control virtual avatars. The software comes with motion-editing features, allowing creators to trim and merge motions, as well as adjust time and position.Theres also a prototype of a spatial capture solution, which is supposed to transform images of real objects and spaces into photorealistic 3D computer-generated assets using Sonys algorithm. One year out from Sonys initial teaser, and details are still pretty slim on what exactly this headset will be capable of, aside from viewing and interacting with 3D objects.Sony is planning to release its Xyn Motion Studio software in late March 2025, but there still isnt any word on the availability or pricing of its headset.

People often follow Probabilities, Deviations and Densities that play a key role in ML modeling. 0 like January 6, 2025Share this postAuthor(s): Chandra Prakash Bathula Originally published on Towards AI. This member-only story is on us. Upgrade to access all of Medium.Photo by Clarisse Croset on UnsplashPart 2: First, make sure you follow these steps, before applying Machine Learning algorithms!We have discussed about basic plotting earlier, now lets dive into some real time insights.We can see the points scatter, but can we work around to get the density of points at a specific region? If there are any, will it be possible for us to know the confidence of the count?To answer these questions we need to delve deep into concepts like Histograms, PDF & CDF.Everything that we have discussed previously is about 2D and 3D. We realized that we could not do visual 4D and developed pair plots.What about 1-D? Why dont we try a 1-D scatter plot?A 1-D scatter plot can be plotted using a single variable on the x-axis to make the y-axis zeroes.By creating data frames based on species in the plot, we have the X-axis as the Petal Length and the Y-axis as nothing.import numpy as npimport matplotlib.pyplot as plt# Assuming iris is a pandas DataFrameiris_setosa = iris.loc[iris["species"] == "setosa"]iris_virginica = iris.loc[iris["species"] == "virginica"]iris_versicolor = iris.loc[iris["species"] == "versicolor"]# Plotting petal lengths for each species using Read the full blog for free on Medium.Join thousands of data leaders on the AI newsletter. Join over 80,000 subscribers and keep up to date with the latest developments in AI. From research to projects and ideas. If you are building an AI startup, an AI-related product, or a service, we invite you to consider becoming asponsor. Published via Towards AITowards AI - Medium Share this post

Optimizing YOLOv8 & Scaling Object Detection on Small Devices 0 like January 6, 2025Share this postAuthor(s): Supriya Rani Originally published on Towards AI. From Quantization Aware Training to Pruning, discover how to make YOLOv8 faster and more efficient for edge devices.This member-only story is on us. Upgrade to access all of Medium.Photo by Museums Victoria on UnsplashImagine running an AI model that can detect objects in real-time on your smartphone or a drone without freezing or draining the battery. Sounds like magic, right? But its the future of object detection, powered by YOLOv8.YOLOv8 is one of the fastest and most accurate models for identifying objects, but its also hefty. When deploying this model on smaller, resource-limited devices like phones, sensors, or embedded systems, you can quickly run into performance issues. Thats where optimization techniques come in.In this article, well explore how tools like Quantization Aware Training (QAT), pruning, and others can transform YOLOv8 into a lean, mean detection machine that works seamlessly on low-resource devices. Whether youre building the next generation of smart cameras or mobile apps, optimizing YOLOv8 is the key to delivering high-performance, real-time object detection.AI generated imageYOLOv8 is the latest release in the well-known You Only Look Once (YOLO) series, and its a game-changer when it comes to speed and accuracy. Its widely used in areas like self-driving cars, surveillance systems, and robotics. But while YOLOv8 is extremely powerful, its also quite resource-intensive. Thats where optimization comes Read the full blog for free on Medium.Join thousands of data leaders on the AI newsletter. Join over 80,000 subscribers and keep up to date with the latest developments in AI. From research to projects and ideas. If you are building an AI startup, an AI-related product, or a service, we invite you to consider becoming asponsor. Published via Towards AITowards AI - Medium Share this post

LatestMachine LearningSmallThinker 3B: A Small Thinking Model Revolutionizing AI Efficiency 0 like January 6, 2025Share this postLast Updated on January 6, 2025 by Editorial TeamAuthor(s): Md Monsur ali Originally published on Towards AI. How SmallThinker 3B delivers big results with minimal resources, making it the perfect choice for edge computing and mobile AI applications.This member-only story is on us. Upgrade to access all of Medium. GitHub | LinkedIn | Medium | Ko-fiPhoto by AuthorIn large language models (LLMs), balancing efficiency and performance remains a critical challenge. With the introduction of the SmallThinker-3B Preview, PowerInfer seeks to address this balance by offering a compact yet potent model tailored for diverse applications. This blog explores the nuances of SmallThinker-3B, its potential use cases, and why it stands out in the competitive LLM landscape.SmallThinker-3B-Preview is a compact yet powerful AI model developed by PowerInfer, designed to deliver high-quality inference while minimizing computational overhead. With 3 billion parameters, it strikes a sweet spot between size and capability, making it ideal for applications where efficiency is critical.Hosted on Hugging Face, the model is easily accessible to developers, researchers, and AI enthusiasts. Whether youre building chatbots, language understanding systems, or creative AI tools, SmallThinker-3B-Preview offers a robust foundation for your projects.Efficiency at ScaleSmallThinker-3B-Preview is optimized for fast and efficient inference. Unlike larger models that require massive computational resources, this model is designed to run smoothly on devices with limited hardware capabilities. This makes it a perfect choice for edge computing, mobile applications, and other scenarios Read the full blog for free on Medium.Join thousands of data leaders on the AI newsletter. Join over 80,000 subscribers and keep up to date with the latest developments in AI. From research to projects and ideas. If you are building an AI startup, an AI-related product, or a service, we invite you to consider becoming asponsor. Published via Towards AITowards AI - Medium Share this post

Is Will Smith about to re-take a famous missed opportunity?As the legend goes, Will Smith was almost cast as Neo in The Matrix back in 1997. He turned it down to make Wild Wild West and Keanu Reeves ended up in the lead role, and the rest, they say, is history. Reeves went on to start in three more Matrix movies directed by the Wachowskis, and Will Smith went on to a have a pretty good career himself. But Smith as Neo was always one of Hollywood's most famous "What if" scenarios.Well, Smith seems to be teasing some kind of Matrix collaboration. On Instagram, Smith posted a video message that outlines the famous story but asks the question, "What would The Matrix have been like with Will Smith as Neo?" and then posts a reply, "Wake up, Will...The Matrix has you..."Sadly, The Hollywood Reporter says Smith won't be starring in the next Matrix movie, directed by Drew Goddard. In fact, nobody seems to know what Smith is teasing here, though we will hopefully find out soon.In the meantime, The Cabin in the Woods director will be writing and directing the new Matrix movie for Warner Bros. Aside from The Animatrix animated film, this is the first mainline Matrix film that won't be directed by the Wachowskis, though Lana Wachoski will serve as executive producer on Goddard's movie.The most recent Matrix movie was The matrix: Resurrections which was released in 2021. Starring Reeves as Neo and Carrie-Anne Moss as Trinity, Resurrections served as something of a conclusion to Neo's story that began in 1997. But given the scale of the universe, there are plenty of other stories that could be told in The Matrix universe.We gave The Matrix Resurrections a 4 in our review, calling it "a bunch of really good ideas stacked together to make a bad - and sometimes ugly - film." Ouch.Matt Kim is IGN's Senior Features Editor.

Dell is offering some great discounts on two of its Alienware m16 R2 gaming laptops: the RTX 4060 model is on sale for $1,299.99 and the RTX 4070 model is on sale for $1,499.99. The m16 R2 is Alienware's most popular gaming laptop, which isn't surprising considering its excellent build quality, top-of-the-line gaming prowess, and reasonable price point. The "R2" is the second generation model that was released in 2024; it's 14% lighter and more compact than its predecessor without compromising on performance.Alienware m16 R2 16" Gaming LaptopAlienware m16 R2 16" Intel Core Ultra 7 155H RTX 4060 Gaming LaptopAlienware m16 R2 16" Intel Core Ultra 7 155H RTX 4070 Gaming LaptopThese two Alienware m16 R2 laptops feature a 16" 2560x1600 QHD+ display, the new 14th gen Intel Core Ultra 7 155H processor, your choice of GeForce RTX 4060 or RTX 4070 GPU, 16GB of DDR5-5600MHz RAM, and a 1TB NVMe SSD. The Intel Core Ultra 7 155H is a Meteor Lake CPU with a Turbo Boost clock speed of 4.8GHz with 16 cores, 22 threads, and a 24MB cache. The Core Ultra 7 is the successor to the Intel Core i7 CPU; it offers similar performance while consuming less power, and thus reducing heat and increasing battery life. If you intend to use this laptop as a gaming rig, there's very little reason to upgrade to the Core Ultra 9 because the vast majority of games doesn't benefit from the additional cores.You absolutely will, however, need a decent amount of GPU power to be able to run games smoothly on the display's bumped-up 2560x1600 resolution. The RTX 4060 GPU with 115W TGP (140W with dynamic boost) would definitely be the bare minimum we'd recommend; its performance lies somewhere between the RTX 3070 and the RTX 3070 Ti, so it's no slouch in terms of gaming performance. The more powerful RTX 4070 with 115W TGP (140W with dynamic boost) is comparable to the RTX 3080 and would fare better at 1440p. Both RTX 40 series GPUs also support DLSS 3.0.Looking for more deals? Check out the best Alienware gaming laptop and PC deals today.Why Should You Trust IGN's Deals Team?IGN's deals team has a combined 30+ years of experience finding the best discounts in gaming, tech, and just about every other category. We don't try to trick our readers into buying things they don't need at prices that aren't worth buying something at. Our ultimate goal is to surface the best possible deals from brands we trust and our editorial team has personal experience with. You can check out our deals standards here for more information on our process, or keep up with the latest deals we find on IGN's Deals account on Twitter.Eric Song is the IGN commerce manager in charge of finding the best gaming and tech deals every day. When Eric isn't hunting for deals for other people at work, he's hunting for deals for himself during his free time.

The US Department of Defense has classified Tencent, a major investor in several big video game companies, as a Chinese military company.The report from Bloomberg states that Tencent alongside battery maker CATL have been added to a list that classifies them as Chinese military companies. The designation itself does not carry any legal consequences such as sanctions, but inclusion on the list is not ideal for companies looking to conduct business in the United States.As a result, while there is no immediate impact to what this designation means for Tencent, this is still a company that owns Riot Games, and has heavy investments in a variety of other game companies including Epic Games, Larian Studios, and FromSoftware. There are also reports that say Tencent is in negotiations with investing more into companies like Ubisoft.PlayIGN's Twenty Questions - Guess the game!IGN's Twenty Questions - Guess the game!To start:...try asking a question that can be answered with a "Yes" or "No".000/250Tencent has told Bloomberg that this designation is "clearly a mistake" and both Tencent and CATL will be pushing to remove their names from the Department of Defense's list. Previously, Chinese phone manufacturer Xiaomi was added to the same list in 2021 and then removed from it a few months later.In a statement to The Verge, Tencent's Danny Marti says that the designation "has no impact on our business" but will still be working to remove themselves from the classification.In a one pager published by the Department of Defense, the US government cites Military Civil Fusion, or MCF, a startegy it says the Chinese government uses to combine civilian research and commercial success to further advance its military, hence why the US government has a list of Chinese companies it suspects as working with the Chinese military, something companies like Tencent and Xiaomi deny.Matt Kim is IGN's Senior Features Editor. You can reach him @lawoftd.

LG unveiled a beautiful new 6K display at CES 2025, dubbed the LG UltraFine 6K Display. It marks the first 6K monitor to incorporate Thunderbolt 5, a feature recently introduced by Apple in the new M4 Pro Mac mini and M4 Pro MacBook Pro models launched in November. Details, including price and release date, are still unknown, but we can learn a few things from LGs teaser.The LG UltraFine 6K display, though still shrouded in technical details, boasts a few crucial specifications. Notably, its a 6K display, though the exact resolution remains undisclosed. Apples Pro Display XDR stands as a comparable 6K display, boasting a native resolution of 6016 x 3384 pixels across a 32-inch panel. This equates to a whopping 20.4 million pixels and an impressive 218 pixels-per-inch. LG also collaborates with Dell to produce a panel for their 6K UltraSharp, which features a resolution of 6144 x 3456.Based on the model number 32U990A, we can confidently deduce that this display is a 32-inch model, matching the dimensions of the Pro Display XDR and the Dell UltraSharp 6K.Another key aspect is the panel technology employed, which is a Nano IPS Black panel. This panel boasts a wide color gamut, spanning 99.5% of Adobe RGB and 98% of DCI-P3. While its definitely not a direct competitor to the Pro Display XDR in terms of color accuracy or contrast ratio, the Nano IPS Black technology offers a significant advantage in terms of cost-effectiveness. By opting for IPS Black instead of miniLED or OLED, LG ensures that the display remains within an accessible price range for consumers.The most pressing question is the refresh rate. Will LG opt for a 60Hz panel, or will the UltraFine 6K offer at least a 120Hz refresh rate? Considering the monitors Thunderbolt 5 connection, I believe a 6K/120Hz option is feasible, but we cant confirm until LG confirms it.A 6016 x 3384 display running at 120Hz would necessitate approximately 79.64 Gb/s of bandwidth. Thunderbolt 5 supports 80Gb/s of bi-directional data and 120Gb/s in one direction with bandwidth boost enabled. However, this is purely speculative, disregarding technologies like Display Stream Compression. Its reasonable to assume that LG would acknowledge the demand for a higher resolution monitor with a higher refresh rate. This would be an ideal companion for Pro Motion-enabled MacBooks. If the price is reasonable, LG could sell many of these units.The other significant factor is adoption. The majority of potential buyers dont own Macs with Thunderbolt 5 capability. Therefore, will the monitor support a lower resolution at a reduced refresh rate, such as 6K60 like the Pro Display XDR? While this seems plausible, well have to wait for more details. However, I believe LG wouldnt want to restrict the market for its new display to Mac owners of the M4 Pro models.The original LG UltraFine 5K Displaythat forehead, though.Finally, lets delve into the design of the LG UltraFine 6K. In a word, its breathtaking. The front panel boasts a striking symmetry, almost devoid of bezels, creating an all-screen illusion. Not a single logo, button, or any other visual element graces the display. This is undoubtedly an improvement over the LG UltraFine 5K Display, which launched back in 2019.The display build materials are intriguing. While Im not certain if its entirely made of metal, the teaser image certainly wants us to believe that. Moreover, it comes equipped with a height-adjustable stand that bears a resemblance to the one found on the Apple Studio Display.Personally, Im very excited to learn more if LG plays its cards right, it could have a bonafide hit on its hands. In the meantime, Id love to hear your thoughts. Feel free to share your opinions in the comments section.Add 9to5Mac to your Google News feed. FTC: We use income earning auto affiliate links. More.Youre reading 9to5Mac experts who break news about Apple and its surrounding ecosystem, day after day. Be sure to check out our homepage for all the latest news, and follow 9to5Mac on Twitter, Facebook, and LinkedIn to stay in the loop. Dont know where to start? Check out our exclusive stories, reviews, how-tos, and subscribe to our YouTube channel

AMD introduced a powerful new laptop chip today, the Ryzen AI Max. The company compared its new chip to Apples M4 line in several benchmarks, but theres a very important detail it left out.New Ryzen AI Max compared to two Apple chips, but not the relevant oneThe new Ryzen AI Max from AMD is a 16-core laptop chip with a heavy emphasis on AI and graphics performance. Its being marketed as an ultra-powerful option for creators, gamers, and more.Naturally, AMD wanted to show off its new chip by comparing it to another option commonly favored by creators: Apples M4 MacBook Pro.The problem is, it left a key M4 configuration out of the test.Paul Alcorn writes at Toms Guide:AMD also included numerous rendering benchmarks of its 16-core flagship against the 12-core Apple MacBook M4 Pro, claiming an up to 86% advantage in a v-ray workload. Naturally, the 14-core M4 Pro, also included in the benchmarks, is more competitive, but AMD still holds a stout lead in the Blender, Corona, and v-ray selection of benchmarks. However, the Ryzen AI Max+ isnt as performant in the multi-threaded Cinebench 2024 test, beating the 12-core M4 Pro by a scant 2%, and trailing the 14-core M4 Pro by 3%.Do you notice whats missing there?AMDs 16-core chip is compared to:Apples 12-core M4 chipand the 14-core M4 ProBut the 16-core M4 Max chip is suspiciously absent from all the benchmarks.M4 Maxs absence from benchmarks makes sensefor AMDDuring AMDs presentation, one presenter touted the Ryzen AI Maxs performance as winning easily against the 12-core [M4], and trading blows with the 14-core [M4 Pro].Apples 12-core M4 only offers a 10-core GPU. With the 14-core M4 Pro, you can get up to a 20-core GPU.Only with the 16-core M4 Max can you get a 40-core GPU.So I suspect that the M4 Max, based on the Ryzens struggles against the M4 Pro, would handily beat AMDs chip in any relevant benchmark.But of course, AMD doesnt want you to know that. Instead, its comparing different classes of chip and claiming victory. What do you make of AMDs comparisons? Do you think theyre fair? Let us know in the comments.Best Mac accessoriesAdd 9to5Mac to your Google News feed. FTC: We use income earning auto affiliate links. More.Youre reading 9to5Mac experts who break news about Apple and its surrounding ecosystem, day after day. Be sure to check out our homepage for all the latest news, and follow 9to5Mac on Twitter, Facebook, and LinkedIn to stay in the loop. Dont know where to start? Check out our exclusive stories, reviews, how-tos, and subscribe to our YouTube channel