images

We have a new winner to generate images with AI. And he is not American, but Chinese

by

Tencent has just launched a new model of AI capable of Generate images from a text prompt. Traditionally the proprietary models have dominated this type of creative task, but the Tencent model has given the surprise and according to various benchmarks is able to generate images better than the rest of competitors, including those of Google and Openai. Hunyuan Image 3.0. This is the name of the new Tencent model, which in the LMARENA classification Of more powerful models for the generation of images from text has managed to overcome Gemini 2.5 Flash image preview (popularly known as Nano Banana), in addition to other proprietary models such as GPT-IMAGE-1, FLUX-1-KONTEXT-MAX or QWEN-IMAGE. The Tencent image model has already managed to overcome its competitors according to LM Arena experts. Blind vote. This LM Arena classification works through a blind voting system in which the users choose their favorite images without knowing what model generated them. And according to this vote, this Tencent model surpassed all its opponents, including Google’s popular “Nano Banana”. Of course: the vote also takes into account long -term results, and the short period that has been available Hunyuan Image 3.0 makes the results qualify as “preliminary” and not definitive. How it works. Those responsible for Tencent explain in the description of the model How they have used a new diffusion architecture that makes use of dual encoders (a multimodal llm and another that better understands characters in different languages) and RLHF optimization (Reinforcement Learning from Human Feedback, which refines the previous result) for the creation of higher quality images. The system makes use in addition to a compression system so that the entire process consumes less resources without loss of quality. Open pesos and commercial license. Hunyuan Image 3.0 is a model that shares Your code in github and that offers A license surprisingly permissive. In fact it is possible to use it for commercial and professional purposes. The price is not entirely economic. Although the model can be tested for free in the Project websitein our tests we could only create an image (10 credits). The platform allows you to buy monthly credits: $ 8 per month allow Buy 500 creditswhich a priori would allow us to create 50 images of 10 credits each. Each would go to $ 0.16, when Nano Banana It has a cost of $ 0.039, four times lower. There are other options to try it, such as Hugging Face “Spaces”. It is also possible to get an API key In Tencent Cloud To use the locally. Gemini continues to win as “editor”. Although the Tencent model is interesting and remarkable, Nano Banana continues to win the game if we consider that it has become a unique substitute for the traditional Photoshop. Many users no longer edit photos but carry one in Gemini and then they tell the AI ​​what changes they want to make in that image. Alibaba, more conversational. Although Hunyuan Image 3.0 can allow something like that – in fact There are demos In this regard – the interface is now more aimed at a single prompt to generate images, not to a “conversation” as GEMINI allows. Another of the protagonists of this land is Alibaba, who with Qwen-Image-Editor adopts the same approach as Google with Gemini and Nano-Banana. In that Alibaba model “you speak” with your image to ask for changes, something that at the moment does not seem that the Tencent model does so directly (although it does not seem difficult for him to get it). But be careful. The differential here is that the generation of images, which seemed to be dominated by proprietary models, can be apparently equally good (or even superior) through open models. One more time The Chinese commitment to that philosophy is remarkable and contrasts with the closed approach and owner of most American companies that develop AI models to generate images and text (or, of course, video). Image | Hunyuan In Xataka | In China they do not conform to create advanced robots: a company has developed a head that gestures like a human

We believed that the J-20 was Beijing’s great furtive jewel. The new images of the J-50 show that there was something else in development

by

“We will strive to achieve the fundamental modernization of national defense and the Armed Forces by 2035, and to make the popular army a World Class Force in all aspects for the middle of this century. ”With these words, Xi Jinping drew in 2017 A military horizon that then sounded ambitious. It could have been in rhetoric, in a gesture designed to dazzle. But in this China the plans are rarely artifice fires: they become state projects, with short and medium term goals conceded into strategies that look forward to decades. And the truth is that a good part of those goals have already begun to materialize. China is not characterized by hurry, although it advances with a speed that surprises. In 2003 he put his first Taikonauta in orbit when the United States and Russia (with the Soviet inheritance in between) accumulated half a century of experience in the space race. Two decades later, Beijing has raised its own orbital infrastructure and a program that does not give brake signals. The same happens on the mainland. In the early 2000s there were no high -speed trains in the country. Today it manages the most extensive network on the planetan example of accelerated transformation. And in defense, which is the focus of this analysis, the jump is equally forceful: Shipyards capable of throwing large -scale ships, aircraft carriers with electromagnetic catapults and last generation combat fighters. J-50, from rumor to the images As Twz collectsthe most clear images have appeared so far from a heavy poacher attributed to Shenyang Aircraft Corporation. In the public conversation it is identified as J-XDs and, unofficially, such as J-50. The material is not verified, but fits with less clear previous shots and does not show obvious indications of manipulation. We see a design without a tail that analysts framed in the next combat generation. J-50 The plane dispenses with tail surfaces, presents tilting pales that appear deflecting down and mounts 2D nozzles with vector thrust and folded bordering. The nose is long, of diamond geometry. Air shots are trapezoidal and adopt a DSI scheme. Under the nose a fairing appears for the electro-optical system and the cabin is a car. Today we do not know if the photographed is a technological demonstrator or a configuration closer to production. There are also no confirmed data on definitive sensor packagepropulsion or calendar. As a context, in parallel there have been advances of the J-20s Biplaza and the J-35 Naval. The focus, in any case, is in this design without a tail that suggests a parallel development to that of J-20. J-20s The photographs provide form, but leave unanswered key issues. It is unknown which propeller equips the plane, what will be its sensor configuration, what communications or electronic warfare systems and what arsenal is planned for him. In any case, EJ-50 becomes the clearest sample that Beijing is willing to experiment with radical concepts in military aviation. A design without a tail, with mobile wounds and vector nozzles, involves entering a land that few have dared to explore. The unknown is how much of what we see corresponds to an isolated prototype and how much is part of a long -distance program. What seems out of doubt is that China wants to position yourself at the forefront of the next generation of poachers. Images | RUPPRECHTDEINO In Xataka | The US has a message for those who turned their backs on F-35: now it will fly alone with its own drone army

Spacex has just published unpublished images of the “Rostized” Starship. A unique perspective of his shock after the toughest reentry

by

Spacex shared new images of the Starship In its tenth test flight, where he managed to complete the reentry and controlled ameter In the ocean despite visible damage in the vehicle. The material spread by Spacex shows the ship at a key moment: the moment of its spareness in the Indian Ocean, dyed of an orange tone after surviving a specially demanding reentry. A key test. The tenth test flight departed on August 26 from Starbase, in Texas, with an impeccable takeoff thanks to the 33 super heavy engines. The separation of stages was also successful and the propeller managed to merit in the ocean, fulfilling its role before the starship continued its trip to space. The milestones in space. Once separated from the propeller, the Starship carried out a complete combustion that placed it in its suborbital trajectory and allowed to validate several key tests. Among them, the deployment of eight Starlink satellite simulators and the second redempted in the history of a Raptor engine in space, two milestones that Spacex considers essential for the development of future missions. Click to see the original publication in x The challenge of the reentry. The reentry was the most critical point of the mission. Spacex had already chained several failed attempts. This, in a way, had questioned the capacity of the vehicle to survive this phase. This time, the ship faced extreme conditions with part of its incomplete thermal shield and the flaps subjected to a deliberate effort. Even with visible damage in the rear skirt, the Starship managed to maintain control and move towards its destination. Spacex’s message. Moments ago, Spacex published a message in which he summed up the scope of what was achieved: “Starship exceeded the reentry with missing tiles intentionally, she completed maneuvers to force her flaps, suffered visible damage in the rear skirt and flaps, and still executed a turn and a landing ignition.” Despite these conditions, the ship managed to preserve sufficient maneuverability to go accurately towards its field of shock in the Indian. The image released next to the statement reinforces that idea. The starship, imposing on the launch platform with all its intact thermal tiles, now appears with a very different appearance: blackened, with a coppery tone that makes it seem almost “roasted” after passing through the atmosphere. Spacex has not explained the exact reason for this change, although on the Internet they have not taken to appear theories. A millimeter closure. The mission culminated with the Starship gently threading in the Indian Ocean, approximately three meters from the planned area. For Spacex, the value of the flight is not that the ship can be reused, but in what has been learned during the test. Each data collected in extreme conditions approaches the company to its goal of developing the first large -sized launcher fully reusable. In Xataka | The “Wow!” Signal signal It was even more powerful than astronomers calculated: half a century later, the mystery is complicated

Nano Banana is not just a great creator of images with AI. It is the greatest danger to Photoshop and company

by

Google has been playing hiding place with what is a great threat to Photoshop, at least in its photo editor facet. For weeks, a mysterious tool called “Nano Banana“He has been appearing anonymously into test platforms, leaving amazed professional users. Google finally confirmed that this experimental model is yours and is integrating it into Gemini. Why is it important. We are facing the beginning of the end of the layers, masks and tools that have defined digital design for 30 years. You change the color of a shirt without distorting the face. You add elements without reinventing the background. In 2 seconds, no 15. For professional uses it is not ready. For first sketches or for domestic use, it is perfect. 10 Google applications that could have triumphed The difference is persistence. Other models generate from scratch with each Promptbut Nano Banana remembers. You can strain about the same image dozens of times. Some examples. Let’s see what can be done with Nano Banana (if you want to know how to try it, We have prepared a guide). Everything has been done are the Prompts that we attach, without more. Original photo: “Create an image with this man but in a Valencian people of the 50s, with clothes, appearance, hairstyle, environment, etc. that are credible for the time.” “Now in the 2050 futuristic Shanghai.” “Now he returns to the original photo and adds this man by his side, posing together for the photo.” “Now back them as two Roman gladiators in the amphitheater.” Let’s go with something else. This Apple announcement in Las Vegas, deployed a few years ago. “Modifies the announcement text to say ‘2×1 in all iPhone models until the end of stock’”. Another different. We give you these two images. “Put this man the red shirt that attached to you.” “Now place him in Mestalla.” “Put a black cap and remove the clock.” All at the blow of Prompt And very fast, more than chatgpt. The money trail. Nano Banana has a cost of $ 0.039 per image of 1024 x 1024 pixels. Adobe has already announced that he will integrate the model in Firefly and Express, as reported Business Insider. His argument (“we offer all the models in one place”) makes sense, although something defensive also sounds. Those who pay dozens of dollars a month for a license to make editions will begin to be renewed. Now it is possible to generate Mockups In minutes, not in days. Some Ecommerce They will stop needing so many photographic sessions. A teacher can create a better adapted diagram than that of publishers. Everything with Prompts of text. In Xataka | Browsers prepare for the most radical transformation in their history. One in which the IA will be Outstanding image | Xataka

How to create images from photos with artificial intelligence

by

Let’s tell you How to create images from photos wearing artificial intelligence. This goes beyond Create dolls with your photos Or similar things, because you can directly make complete images from your photos, change your appearance, and even add things. Let’s start by giving you a series of prior tips to take into accountwith several things you will need to know. Blabla Before starting, some tips Before you start, the first thing you have to do is know exactly what you want. By this I mean that you should know if you simply create a totally different image from a photo, or if you want to transform this photo into a different one. Some of the things you are going to do are the following: Convert the photo into a drawing with various types of styles available. Create an image using the photo as a gestural reference, but changing all the content. Alter the photo to change your format or remove the background. Maybe even eliminate elements This may seem nonsense, but depending on what you want there are tools that will be more useful than others. Another thing to keep in mind is that you should try Use good quality images As a reference, with good resolution and good lighting, trying that there are no blurred objects. I also recommend you experiment In the event that you want to reimagine the image to create a drawing from it. In this article We give you many styles ideas to edit your photos. It is also important take into account the limits of use of each tool. There are some that are free but that only allow you to use them once or twice before paying, and there are more flexible ones, but with less functions. In general, working with photos is usually a payment function. The best free tools Now let’s tell you The best free services and tools To create images from photos using artificial intelligence. Chatgpt: Chatgpt being the AI that carries the front in terms of making images from photos, and is the one that offers the best results. However, in the free version the process can take a long time. Link: chatgpt.com Microsoft Copilot: COPILOT It is also able to offer very good results, since it is using the same technology as Chatgpt, but slightly modified. They are somewhat different results, but equally good. Link: Copilot.microsoft.com Fotor: It has several artificial intelligence tools with which you can eliminate funds, change sizes or make modifications. Quite versatile. Link: fotor.com/photo-editor-app Grok: X’s AI can do something similar, but the results are quite bad. It is good to generate images from nothing, but bad to create them from photos. Link: Grok.com Google Gemini: We mention a Gemini for being another of the great generative, but for the moment It does not allow images based on photosthat is not to serve you. Link: Gemini.google.com IdeOGram: IdeOGram It also offers interesting results. Unfortunately you can only do it if you are a payment user, with the modality of 16 euros per month. Link: ideogram.ai Freepik: This model is a specialist in making compositions based on faces. However, it is also necessary to use your payment plan to create images using a photo as a reference. Link: Freepik.com Picart: You can add effects to a photo and make editions, such as deleting its background or changing colors. Link: picart.com Vondy: It specifically serves to create a cartoon from the photo you upload, modifying factions of the face to make it fun. Link: Vondy.com Create images from photos with chatgpt Chatgpt is the most capable tool To create images from artificial intelligence. You have many possibilities, and the way to use it is using a Prompt written next to the photo you upload. Every time you repeat the prompt with the same photo you will have a different result. To create an image from a photo you must attach the photo next to a prompt To say something like “reimagine this photo with the style of ..”. If in the end you add an image or cartoon style, it will turn that photo into a cartoon of what you have asked. You can also use it to make cartoons from photos. You can also create images using photos as a reference For the position you want to give. For example, I have uploaded a photo of my cat and asked to draw a lion with the same position, and Chatgpt will do it. You can also specify that the background changes. Another curiosity you can do is take the face of a photo and put it in a drawing that you have created by artificial intelligence. This can be used for jokes above all, because the difference will be noticed. Of course, you can also ask that Use the face as a reference To create the character. Finally, you can also Remove the background to an image To maintain what appears in the foreground. Thus, you will create a PNG with a transparent background that you can use on other occasions. Of course, specify well what you want to keep in the foreground, for example, to say that it is a person or an animal, to take it into account. And the other alternatives? Currently, the AIs that best work with your photos to create images from them are Chatgpt and Co -ilotboth based on the same Openai technology. The rest of the alternatives are simply far from the results that these two offer. Even some big names of AI, such as Google Gemini, cannot even edit images from photos. You can try to use the alternatives from the list, some of them are very useful for specific editions. However, when creating images from a photo they are quite bad and lower. Of course, remember that the evolution of AI is being dizzy, so All this could change in a few months. Therefore, pay attention to how tools evolve over time. In … Read more

We have tried to use AI to verify if the images of the war between Iran and Israel were made with AI. It has been a disaster

by

False images circulate on social networks is no novelty, but in the era of AI It is increasingly difficult to detect them And they have even made Let’s distrust real photos. That photos taken with AI are viral when an important event is the new normality is; We saw it After the blackout And also with him Conflict between Israel and Iran. It is clear to us that AI is very good to generate false images, but what if we use it precisely for the contrary? Not so much. The photos in question. At the same time as a Half Iranian published The news that Iran had demolished an Israeli F-35, two images began to run like gunpowder on social networks, although We soon knew they were false. One of them, that of the star and surrounded by curious F35, is especially striking. To begin with, the proportions make no sense: the plane seems giant when actually measures 16 meters and people are larger than buildings. And that not to mention that the damage of the plane is minimal to have been shot down. These images were not the only ones generated by AI that circulated in the first days of the conflict. Several videos such as this one of a huge Iranian missile that it seems quite real until we see that the water brand has been left to see that it is made with the, or East of a tel aviv shattered. AI is terrible doing FACT-CHECK. Means dedicated to FACT-CHECKING as Damn already denied These and other images created with AI in the context of the conflict between Iran and Israel. However, there were users who tried to resort to AI tools to check the authenticity of the images and obtained quite confusing answers. Is What happened in X with Grok. An analysis of more than 130,000 posts revealed that Musk’s AI was not able to detect some false images or identify the sources from which they came. The community notes written by the users themselves were much more reliable. We have tried it. To check the AI ​​capacities, we have used the image of the disproportionate F-35 and have asked several AI tools. This is what they have answered: Chatgpt: The Openai tool begins “This image does not seem real” and then proceeds to make an analysis of the proportion of the plane, which correctly identifies as a F35, and states that the damage does not seem coherent. Perplexity: Like Chatgpt, he tells us that proportions, perspective, and airplane damage and other details suggest that the photo has been digitally manipulated. Gemini: It tells us that the image is real, but that it is not an attack in combat, but of a clash with birds that happened in Israel in 2017. When we answer that the sources show us, it happens several links to the news, but in none of them the image appears. After a while sending us confusing information, he ends up recognizing that he was wrong and apologizes for “the serious mistake.” Claude: It is the only one that states with forcefulness that the image is not real and gives us the exact context of what has happened “this is one of the many false images that have circulated as part of misinformation campaigns during the conflict between Israel and Iran.” The reliability, pending subject. In our test, Gemini has completely invented the answer, while Chatgpt and Perplexity succeed, although they do not get wet. Claude is the only one that gives us all the information and hits fully. Although language models have improved a lot in a short time, Many answers continue to be invented Despite having access to the Internet and searches. Undoubtedly, reliability is the pending subject of generative AI and where more improvement margin has. Images | 404 average In Xataka | Chatgpt guide: 22 functions and things you can do to squeeze this artificial intelligence to the maximum

There are almost no images of the conflict between Israel and Iran. That is why we are starting to generate them with ia

by

At the user level, the popularization of artificial intelligence is leading to everyday uses (of cheat until helping us Organize the day) already debates that affect us all (ecological, labor and Ethics). But its use is also in the hands of governments and, when they conflict, artificial intelligence becomes a valuable weapon of soft war and propaganda on a global scale. He Wask between Iran and Israel He is leaving us a few examples. The images about these lines are not real. They are generated by AI. When the Gulf War in 1991 It was said that it would be the first war that we would see live, and now we are prolonging that same idea but taken to the extreme … and seasoned with manipulation of information through artificial intelligences. Images and videos disseminated in social networks and messaging apps, where we can see live bombings, but where not everything is real. These are images of some assumptions F-35 demolished by the Iranian army, data backed by the news of the country. Israel denied it quickly. And you just have to take a look at the second photo to see that it is false: what absurd size does that hunt have? And the buildings in the lower corner are smaller than people. That did not prevent thousands of people from spreading the images in their accounts, giving them by legitimate. And also videos. The arrival of the potent I see 3 Google has helped generate videos in line with these photos. The Tehran Times, means of Iranian communication, disseminates videos where it is presumed to gigantic missiles that do not exist. How do we know? I have left the seeing water mark in the video. Something similar happened with a supposed bombardment in Tel Avivwhose images are easy to identify as generated by AI. The person responsible for all of them is the account 3AMelyonnwhich is defined as “artificial intelligence resistance”, and whose videos are also spread on Telegram. With the permission of the high positions. These actions are not always a matter of disinformation in networks. Iran’s supreme leader, Sayyid Ali Khamenei, tweet images of Iranian missiles to Israel generated by chatgptand the Israeli defense minister generates Genuine visual garbage (which The Anglo -Saxons call ‘Ai Slop’) of propaganda type of which Facebook and other networks crowd. It is not strange that the leaders of both sides join the iconographic war: it is a way not only to spread harmful propaganda for the other side, but to supply the informative limitations suffered by citizens. Please, only AI. In fact, there is another reason for the dissemination of propaganda through AI to be backed by the respective governments of Iran and Israel: when citizens spread images of killings in the cities, it was feared that with these videos they would give strategic clues to the other side. AI is a way to ensure that no one takes advantage of those videos. According to 404 mediaIsrael requested from its citizens in a statement by social networks: “The enemy examines these documents to improve their objectives. Be responsible, do not share locations on the network.” AI thus becomes a tool to share propaganda that is not strategically sensitive. How to identify artificial intelligence. The companies that create the generative tools are responsible for marking the results that their IASs yield. In addition to easily visible and editable water marks, Google AI includes an invisible, Synthidbut it has a problem set: you have to download the video that may appear, for example, on social networks, and then upload it to the Synthid platform. They are minutes, perhaps lost hours, during which perhaps the false video has reached millions of people. A very valuable time in terms of propaganda. Header | 404Media In Xataka | Welcome to the “mcdonaldization” of AI: we have cheap chatbots for mass, and ultra -premium products for the elite

Satellite images have revealed what happened to one of Russia’s biggest arsenals. Now we understand Moscow’s silence

by

On April 22 the satellites began to point out A point on the planeta change only perceptible through the images from space offered a first track of what was happening about 60 kilometers from Moscow. Despite the weather conditions of that day and the low resolution of the optical data captured by the Sentinel-2 satellite From the European Space Agency, the damages were clearly visible. An explosion had “burst” the 51st arsenal of the main missile and artillery direction of the Russian Ministry of Defense. Total devastation of Arsenal. Visual confirmation was reinforced by radar images Synthetic opening (SAR) capable of penetrating clouds and smoke, which showed significant structural alterations in the complex nucleus. The comparison between images Taken on April 14 and 23, it indicated that at least 30 buildings destined to storage of ammunition had been completely destroyed. Explosions, evacuations and blackouts. The day after the explosion, the secondary detonations They still continuedunderlining the magnitude of the stored material. The strength of the outbreak forced Evacuate eight nearby townswhile 37 settlements were left without gas supply. The most remote evacuated town was 4.5 kilometers from Arsenal. NASA fire monitoring system data also confirmed the existence of multiple igneous foci Within the perimeter, coinciding with the analysis of the intelligence expert (OSINT) MT Anderson, who used additional filters to detect heat points and Confirm destruction Massive infrastructure. A strategic arsenal. Then the magnitude of what happened began to be known. He 51st Arsenal Grau It was not simply a deposit of ammunition. As one of the Eight main arsenals that still operated in the European part of Russia, its function was key both in the distribution and in the logistics maintenance of the Moscow weapons. Three of those eight arsenals had already been destroyed for 2024, which turned this loss into a considerable strategic blow for the Kremlin military supply chain. Arsenal was designed to house Up to 264,000 tons of explosive material. Among the remains found after the explosion were identified 107 mm rockets for Multiple type 63 rocket Chinese manufacturing, many of which were recorded spread around local residents, suggesting that part of the material was stored outpatient and had recently been delivered. The catastrophe, or the attacknot only compromised Russian logistics operability in the Ukrainian conflict, but raised (once again) serious doubts about the security of its own arsenal in times of war. Images of the British report with the before and after the explosions A self -inflicted blow. Now, and after A study Of all the images and confidential information of the intelligence of the United Kingdom Ministry of Defense, it has been confirmed that the cause of the incident was not “external”, but a combination of bad practices in the management of armament and a negligent storage management by Russia. British research, in fact, is reinforced by the declaration of the Russian Defense Ministry itself, which, in silence from the incident without offering more data, there were attributed the disaster to the “violation of security requirements” in the manipulation of explosive materials. For the United Kingdom, the event is not an isolated case, but the reflection of a prolonged and documented trend of “Russian ineptitude in the treatment of its own ammunition”, although that yes, in this case it represents the greatest loss of self -inflicted arsenal since the beginning of the large -scale war in Ukraine. Strategic installation We already said it before. The affected deposit was a key installation for the war supply of the Kremlin on the Ukrainian front and, according to figures from the Ukrainian authorities cited by the United Kingdomhosted around hundreds of thousands of tons of ammunition, including ballistic missiles, projectiles thrown from air and anti -aircraft systems. Satellite images verified by the insider medium They also revealed that more than a square kilometer of the complex was affected by the detonations, which suggests that massive and prolonged destruction, with multiple fires and a chain of secondary explosions that, According to disseminated videos In social networks, they even reached nearby civil areas. Error pattern. In addition, it is not the first time that the arsenal of the 51st Grau suffers incidents of this type. Insider told That in June 2022, Russian state media reported a spontaneous explosion during loading and unloading operations that cost four people. The pattern is consistent with British complaint: A continuous chain of operational errors and insufficient security measures that make critical facilities into vulnerable points within the Russian military apparatus. The lack of technical discipline and effective prevention protocols has not only generated large material losses, but also has compromised the safety of populated areas in times of war. Consequences. If you want also, the incident gives wings to the rhetoric of the West. The impact of this catastrophe transcends the material. The destruction of one of the main deposits of Russian ammunition not only weakens the immediate logistics capabilities of Moscow in its offensive against Ukraine, but also reinforces an idea increasingly sustained Among the “alidos”: that of a corroded military power for structural failures, operational improvisation and a dangerous carefree for the most basic security standards. Seen thus, in full prolonged war and with its supply lines under pressure, losing tens of thousands of tons of armament due to internal negligence constitutes a defeat with several readings. Image | Maxar In Xataka | Russia launched its fearsome nuclear missile Satan II last week, the “Invincible Weapon” of Putin. It was regular In Xataka | The US has detected an object in space with strange behavior. The source that released it has also located: Russia

Freepik has just launched an Open Source model with a groundbreaking characteristic: licensed images

by

The Spanish Startup Freepik is already one of the absolute referents in the segment of artificial intelligence, but its last launch is especially significant. It is a new model of its own generative called F lite which is a new blow to the table for several reasons, but especially for an especially striking. Licensed images. Freepik, in collaboration with Fal.AI, has presented F lite, a model of generative text in image that stands out especially for being trained “exclusively with high quality images, with legal support and copyright protected” thanks to an important detail. They come from the Freepik library. Inspired (partly) in Deepseek. In Xataka we have talked to Omar Pera (@ompemi), Product responsible in Freepik. He has explained to us how the Deepseek Chinese model served as inspiration by demonstrating that it was possible to create a small but very capable model with much less resources and data. F lite is trained with “only 80 million images, compared to more than one billion usual images” in models of the generative images of the competition. Open Source. As explained in the technical report that accompanies the launch, F Lite is also An Open Source model of 10 million parameters-diminuto if we compare it with the 1.76 trillion estimated GPT-4-which has been trained for two months in 64 GPUS NVIDIA H100. A sample of the result obtained with F lite via Fal.AI. The car does not resemble Renault 5, true, but at least at the level of photorealism the result is really decent. Small but solvent. As explained Iván de Prado (@ivanprado), one of the top people responsible for its development, F lite is a decent model to generate certain types of images despite its small size. It also has its limitations, and can show anatomical defects or not give good results in complex or text rendering. The model is available in two versions in Hugging Face (regular, Texture) and in Fal (regular, Texture). It is also possible Download the github and use it at home, for example via comfyui. Is just the beginning. The launch of this model raises the beginning of an especially striking project that could gradually make it an alternative that rival the most ambitious models, but always supported by that argument of being trained with licensed images and being Open Source. More options for users. The model is not part of the moment of the offer of models available on the Freepik web platform, yes, and does not compete directly with models such as FluxMystic or image 3. “The bet and strategy does not change,” Pear pointed out, and consists in “optimizing the user offer and offering the best technology to solve the problem” and the need of each user. Demands everywhere. We have been seeing how artificial intelligence companies make indiscriminate use of content protected by copyright to train their models. They usually do it without a license and without having permission to use those works, and that has caused numerous Copyright rape demands, especially in cases Like OpenAi. Freepik Enterprise. This announcement comes almost at the same time as its new business offer, called Freepik Enterprise. Omar Pera confirmed that the initial reception of F Lite has been remarkable since this new service. In this area it is precisely where using a model like this is especially interesting, because “companies are covered” when using a model trained with licensed images. A by adobe and other rivals. Pera also pointed us out in Freepik They do not compete With images, “we are going for professional use cases of marketing or design.” They are not compared to a Getty or a Shuttersock, and more “with the creative tools of Adobe, with Leonardo (part of Canva from a year ago) or with the professional services of Midjourney. “ Image | Freepik In Xataka | All the great AI have ignored the laws of Copyright. The amazing thing is that there is still no consequences

After the blackout, false images of Spain and Portugal circulated from space. Now we have the real photos

by

55 million people ran out of electricity in Spain and Portugal on April 28, but Total darkness images They circulated the next day, like the one above, they were false. The Balearic Islands did not suffer the blackout, and a good part of the Peninsula already had light when the night fell. Now the European Space Agency has compiled the real images captured from space. The trigger for this collection work were, in fact, the false images. Alejandro Sánchez de Miguel, researcher at the Institute of Astrophysics of Andalusia linked to ESA light pollution projects, He saw the photos that were circulating And he decided that the real ones had to be published. Three NASA satellites equipped with Night observation technology (Suomi-NPP, NOAA-20 and NOAA-21) spent a couple of times on the peninsula that night. Their Six images They tell us a nuanced story, and very different from viral montages, how the Peninsula finished illuminating. The blackout at 03:12, 03:36, 04:30, 04:54 and 05:18 While areas like Madrid had regained light around 22:00 on the 28th, other regions, especially in the souththey followed in the dark until Well enter the morning On the 29th, the almost complete recovery, visible in the last passes of the satellites, arrived at 5:00. The night was clear in almost the entire territory. Dark spots in France or the Portuguese coast are not due to supply cuts, but that the satellite did not go through that concrete region in the waterfall. In green, the areas that still had no light. Blank, the ones In these contrasted images, areas without light can be visualized more easily in green, while areas with electricity supply appear blank. The provinces of Almería and Granada They were the ones that took the closest to illuminate. There are also green areas in Castilla-La Mancha and dispersed regions of Levante, the Sierra Morena and the Campo de Gibraltar. The blackout in Andalusia seen by NASA Earth Observatory While satellites help quickly evaluate the scope and progression of light cuts, the blackouts itself offer space agencies the possibility of Study light pollution and its impact on the observation of heaven or In circadian rhythms of people. In areas such as Almería, the light pollution It was reduced between 70 and 80% during the blackout of April 28. But the light was restored, the stars turned off again and Pilas radio He went back to the drawer. Images | NASA, that In Xataka | ESA has launched the world’s first satellite equipped with Band Radar P. The goal: see through forests

Log In

Forgot password?

Forgot password?

Enter your account data and we will send you a link to reset your password.

Your password reset link appears to be invalid or expired.

Log in

Privacy Policy

Add to Collection

No Collections

Here you'll find all collections you've created before.