Thursday, January 15, 2026
ad
Home Blog Page 54

CVPR 2023 Announced Best Paper Awards

CVPR 2023 Best Paper Awards
Image Source: RSIP Vision

The Computer Vision and Pattern Recognition (CVPR) Conference Awards Committee announced the winners of the highly anticipated Best Paper Awards for 2023. CVPR 2023 is hosted by the IEEE Computer Society (CS) and the Computer Vision Foundation (CVF). CVPR is the leading conference for new research in computer vision, artificial intelligence (AI), machine learning (ML), augmented and virtual reality (AR/VR), and deep learning (DL). 

The CVPR 2023 received an impressive 9,155 paper submissions, with each submission undergoing a thorough evaluation by a minimum of 3 reviews from a pool of 6,625 reviewers. Out of this rigorous process, 2,359 papers were accepted, resulting in a 25.8% acceptance rate. After careful evaluation the CVPR 2023 Awards Committee handpicked 12 outstanding papers as the Best Paper winners of the year.

The announcement was made during a special keynote session on June 21, 2023. The winners of the CVPR 2023 Best Paper Awards are as follows:

Read More: YouTube to Soon Add AI-powered Dubbing Service 

In addition to the Best Paper Awards, IEEE CS also announced the recipients of the Technical Community on Pattern Analysis and Machine Intelligence (TCPAMI) Awards at the conference. 

  • The Longuet-Higgins Prize was awarded to the 2013 paper “Online Object Tracking: A Benchmark” by Yi Wu, Jongwoo Lim, and Ming-Hsuan Yang. 
  • The Young Researcher Award was given to Christoph Feichtenhofer and Judy Hoffman.
  • The Thomas Huang Memorial Prize was awarded to Alyosha Efros.

These awards not only celebrate the lasting impact of research presented at CVPR but also highlight the continued advancements made by the computer vision community.

Advertisement

David Ha, Stability AI Head of Research, Resigned!

David Ha Stability AI Head of Research Resigned
Image Source: Mashable

Stability AI, the highly anticipated artificial intelligence startup, is undergoing a significant leadership shake-up as two top executives, including the head of research and the chief operating officer, have recently resigned. The departures have sparked curiosity and raised questions about the company’s future direction.

David Ha, who served as the head of research for the London-based startup, stepped down earlier this month, and Chief Operating Officer Ren Ito also parted ways in June. While David Ha has chosen to remain tight-lipped about his departure, Ren Ito has yet to respond to requests for comment, leaving industry insiders speculating about the circumstances surrounding these sudden exits.

In response to inquiries, Stability AI spokesperson Motez Bishara issued an official statement shedding light on the matter. Bishara confirmed Ha’s departure, citing personal reasons for his decision. Regarding Ito, it was revealed that he left to pursue other interests. Bishara expressed gratitude for their contributions to the company and extended well wishes to both individuals.

Read More: Stability AI Announces Release of Open-source StableStudio

The founder and CEO, Emad Mostaque, faced allegations of a “history of exaggeration” in an article by Forbes, which he denied. 

Stability AI, previously valued at an impressive $1 billion by investors, gained recognition for its groundbreaking development of Stable Diffusion, an open-source image generator publicly released in August 2022. Runway and University of Munich researchers have collaborated in 2021 to create the initial version of Stable Diffusion. Stability AI, later supported the project by funding extensive data training and development. Stable Diffusion swiftly established itself as a strong competitor to OpenAI’s Dall-E, driving Stability AI to develop and unveil improved versions of the tool.

With a strong commitment to advancing AI and a track record of innovation, Stability AI is poised to overcome challenges and make significant contributions. Under evolving leadership, the company’s future holds intrigue and anticipation in the tech community.

Advertisement

Microsoft Acquired Bethesda After Knowing Starfield Would be Exclusive to PlayStation

Microsoft acquired Bethesda knowing Starfield exclusive PlayStation
Image Credits: Success Stories

In March, ZeniMax Media, the parent company of Bethesda Softworks, was acquired by Microsoft. This arrangement covered Bethesda subsidiary studios like Arkane, the creator of Dishonoured, MachineGames, the creator of Wolfenstein, id Software, the creator of Doom, and Tango Gameworks, the creator of The Evil Within.  

One of the primary drivers behind Microsoft’s acquisition of ZeniMax, the parent firm of Bethesda, has been revealed by the Xbox head of Microsoft. It is the “possible Starfield PlayStation exclusivity.” Phil Spencer, CEO of Microsoft Gaming, revealed today at the FTC vs. Microsoft hearing that Sony frequently pays rivals to “skip our platform” and that Microsoft believed it needed to buy Bethesda to compete.

“When we acquired ZeniMax one of the impetus for that is that Sony had made a deal for Deathloop and Ghostwire to pay Bethesda to not ship those games on Xbox. So the discussion about Starfield when we heard that Starfield was potentially also going to end up skipping Xbox, we can’t be in a position as a third-place console where we fall further behind on our content ownership so we’ve had to secure content to remain viable in the business,” said Spencer. 

Read More: Vimeo Introduces AI-Powered Script Generator And Text-Based Video Editor

ZeniMax Media, the parent company of Elder Scrolls and Fallout studio Bethesda Softworks, was purchased by Microsoft for $7.5 billion. Microsoft promised Xbox and PC exclusives at the conclusion of the agreement, and so far Redfall has been released, with Starfield scheduled to launch on September 6th. The forthcoming Indiana Jones game from Bethesda is only available on Xbox and PC.

Today, a significant portion of Spencer’s testimony focused on presenting Sony as a combative and aggressive rival. Spencer claims that the only way to compete with Sony is by buying up Bethesda and attempting to acquire Activision Blizzard. The estimated value of the planned Activision Blizzard acquisition is $68.7 billion.

Advertisement

Databricks Acquires OpenAI Competitor MosaicML for Whopping $1.3 Billion

Databricks acquires MosaicML for $1.3 Billion
Image Credits: AD

A 2-year-old generative artificial intelligence startup called MosaicML will be acquired by Databricks, a $38 billion company, in a deal worth $1.3 billion in stock, with the retention benefits of the MosaicML team. According to the official website, the team consisted a total of 60+ members, combining the staff of offices in San Francisco, New York, Palo Alto & San Diego.

The whole MosaicML team, including its top-tier research team, is anticipated to join Databricks following the acquisition. According to Databricks, the partnership with MosaicML would make generative AI available to all types of businesses, allowing them to create, control, and secure generative AI models using their own data.

Naveen Rao, a former executive at Intel and the founder of Nervana and Hanlin Tang established MosaicML in 2021. From investors including DCVC, AME Cloud Ventures, Lux Capital, Lux, Frontline, Atlas, Playground Global, and Samsung Next, MosaicML had raised $64 million.

Read More: Vimeo Introduces AI-Powered Script Generator And Text-Based Video Editor

Two AI solutions from MosaicML were unveiled last month in an effort to undercut market leaders like OpenAI on price. Numerous generative AI services offered by MosaicML compete more directly with firms like OpenAI or Anthropic.

Modern MPT large language models (LLMs) are a trademark of MosaicML. With more than 3.3 million downloads of MPT-7B and the most recent launch of MPT-30B, MosaicML has demonstrated how businesses can easily create and train their own cutting-edge models using their data in a time- and money-efficient manner. These AI services are available as much as 15 times less expensive than its rivals, according to MosaicML.

Advertisement

Meta Partners with CAIT to Train 1 Million Traders in WhatsApp Business App

Meta partners CAIT train 1 million traders WhatsApp Business App
Image Credits: AD

On June 26, Facebook parent company Meta announced a collaboration with the Confederation of All India Traders (CAIT) to train one million traders in the WhatsApp Business App over the course of the following three years. The company stated that this programme will include thousands of trade groups affiliated with CAIT across the nation.

Praveen Khandelwal, National Secretary General of CAIT said, “The five essential foundations for India’s trading community are networking, logistics, communication, technology, and the internet. With businesses of all sizes using WhatsApp to communicate, coordinate, and interact with their consumers successfully, there is an urgent need to equip them with the skills needed to promote company growth.”

Additionally, the business has announced the opening of Meta Small Business Academy, a programme that trains up-and-coming business owners and marketers in digital marketing techniques so they may succeed on the Meta’s family of applications, which includes Facebook, Instagram, and WhatsApp, while earning an award certification in the process.

Read More: Vimeo Introduces AI-Powered Script Generator And Text-Based Video Editor

In order to serve Micro, Small, and Medium Enterprises (MSMEs) across the nation, the training module and examination will be made available in six Indian languages in addition to English, including Hindi, Marathi, Bengali, Kannada, Tamil, and Telugu, the company stated in a statement.

As a strategic partner for its MSME skilling programmes, Federation of Indian Chambers of Commerce & Industry (FICCI) is listed by Meta as having created numerous programmes over the past few years to meet the specific needs of the nation’s MSMEs.

FICCI and Meta helped 500,000 women-owned small and medium-sized businesses (SMBs) last year by teaching them digital marketing skills. “In the future, we look forward to partnering with Meta to build a strong ecosystem for small businesses across the country, especially women-led businesses,” Khandelwal said.

Advertisement

Meta Launches New VR Subscription Service Meta Quest+

Meta VR Subscription Service Meta Quest+
Image Credits: Stock Images

To help shape a budding yet highly-invested market, Meta Platforms on Monday introduced Meta Quest+, a subscription-based service for its virtual reality (VR) devices.

CEO Mark Zuckerberg announced that the Meta Quest+ membership will be available starting on Monday for $7.99 (about ₹655) per month or $59.99 (roughly ₹4,920) per year for its Quest 2, Pro, and soon for Quest 3. Since its major bets on the metaverse failed to pay off, Meta lowered the price of their headsets in March.

According to Meta’s website, Quest 2 is being sold for $299.99 (almost ₹24,600), and the company’s flagship VR headset, Meta Quest Pro, is currently available at $999.99 (nearly ₹82,000), down from its launch price of $1,499.99 (nearly ₹1,23,000).

Read More: Vimeo Introduces AI-Powered Script Generator And Text-Based Video Editor

Although virtual reality headsets have become the “next big thing,” their adoption has been restricted to the gaming industry despite the devices’ advanced capabilities. 

Every month, members of the Meta Quest+ subscription service will have access to two carefully chosen VR titles. Additionally, if users sign up before July 31 with Meta, they may get their first month for just $1 or ₹82, which is a limited-time offer. Only new subscribers qualify for the promotion. 

Meta said in an official statement, “At a monthly value of up to $60, Meta Quest+ combines affordability with the convenience of a curated experience. We’re excited to give gamers on Meta Quest a new way to explore all their headsets have to offer.”

Advertisement

Sony Music Hires Geoff Taylor as Executive VP of AI

Sony Music Hires Geoff Taylor Executive VP AI
Image Credits; Sony Music

According to an internal memo, Sony Music has appointed former BPI CEO Geoff Taylor to the position of executive vice president of AI, possibly the first senior executive position in the music business dedicated to AI. 

Taylor, a Member of the Order of the British Empire and former CEO of the British Phonographic Industry (BPI) trade association for 15 years, will report to Sony Music Entertainment COO Kevin Kelleher in his new position.

Taylor will be in charge of overseeing the company’s AI-related business initiatives. He will also be looking after the business and legal affairs, global digital business, and other divisions, according to what the memo states.

Read More: Vimeo Introduces AI-Powered Script Generator And Text-Based Video Editor

“Geoff brings to our company decades of music industry experience,” Kelleher said. Prior to joining the BPI, Geoff was General Counsel and Executive Vice-President at the global recorded music trade body, the IFPI from 2005-2007.

Since the introduction of streaming, nothing has had as much of an impact on the industry as AI in recent months. In contrast to how it responded to illicit downloading and file-sharing at the beginning of this century, the industry is now acting considerably more proactively and is developing standards and guidelines for authorizing and paying artists and rights holders. Taylor’s appointment is unquestionably representative of that strategy.

Recently, major changes were revealed by The Recording Academy for the biggest award show in music. The Grammys will now accept entries for music created by artificial intelligence, although it will not be considered for nomination or for award representation.

Advertisement

Anand Mahindra Defends Sam Altman’s comments on Indian ChatGPT

Anand Mahindra defends Sam Altman's comments Indian ChatGPT
Image Credits: AD

Sam Altman, the CEO of OpenAI, traveled to India earlier this month to meet with the country’s prime minister, Narendra Modi, and other industry leaders to discuss the future of AI. 

However, the creator of ChatGPT also came under fire for his remarks about India’s capacity to produce something akin to the ground-breaking AI chatbot. Later, Altman explained his position, claiming that the question posed to him was inappropriate. Anand Mahindra has since added to his defense of the OpenAI CEO, claiming that he was misunderstood.

Anand Mahindra said in a tweet that he recently visited the Tech Handshake event at the White House and met OpenAI CEO Sam Altman there. He claimed that Altman’s remarks about India developing ChatGPT-like technologies were misinterpreted, and he added that he has no doubts about India’s potential. He also made reference to CP Gurnani, CEO of Tech Mahindra, who tweeted that he had “accepted” Sam Altman’s challenge.

Read More: Vimeo Introduces AI-Powered Script Generator And Text-Based Video Editor

“The Tech Handshake meeting this morning at the White House was refreshingly frank, thanks to the direction of Gina Raimondo. My optimism about closer technology cooperation is because mutual benefit is now involved rather than just a one-way request from India. On the sidelines of the meeting, I caught up with Sam Altman about the ‘challenge’ that CP Gurnani had accepted. Sam reiterated that he’d been misunderstood. He’s far from skeptical about Indian abilities,” Anand Mahindra tweeted.

When Tech Mahindra CEO CP Gurnani interpreted Altman’s statement as a “challenge,” the conversation attracted even more attention. “OpenAI founder Sam Altman said it’s pretty hopeless for Indian companies to try and compete with them,” he commented on Twitter after sharing the video. “CHALLENGE ACCEPTED,” he added.

Advertisement

LinkedIn And UC Berkeley Introduces New Method To Detect AI-Generated Profile Photos

LinkedIn UC Berkeley New Method Detect AI-Generated Profile Photos
Image Credits: LinkedIn Blog

In a collaboration between UC Berkeley, LinkedIn has come up with cutting-edge detection techniques that correctly identify artificially created profile pictures 99.6% of the time. Only 1% of real photos are mistakenly identified as fake.

Two forensic techniques can be applied to look into this matter, according to the blog. Methods based on hypotheses can identify peculiarities in faces created artificially. Learning blatant semantic outliers assists this method. Second, data-driven methods like machine learning can be used to tell natural faces apart from CGI ones. 

The suggested work uses a hybrid methodology that involves first identifying a distinctive geometric feature in computer-generated faces and then using data-driven techniques to quantify and detect it. This approach uses a lightweight classifier that can be trained fast, and it only needs a small number of synthetic faces. 41,500 synthetic faces were created using five different synthesis engines, and 100,000 genuine LinkedIn profile photographs were used as extra data.

Read More: Vimeo Introduces AI-Powered Script Generator And Text-Based Video Editor

Researchers compared an average of 400 of each of actual (publicly available) LinkedIn profile photographs with artificially generated (StyleGAN2) faces to see how they differed. The majority of real profile photographs are essentially stock headshots because people’s actual photos vary so much from one another. 

The average StyleGAN face, in contrast, has very distinct features and keen eyes. This is because of the standardized interocular distance and ocular position of StyleGAN faces. While StyleGAN faces are normally created from the neck up, real profile shots frequently concentrate on the upper body and shoulders. Researchers also sought to focus on the visual similarities and discrepancies that can be observed within and among distinct social groups. 

Advertisement

Chennai Police Installs AI-powered Security System of 5,250 CCTV Cameras

Chennai police installs AI-powered security system 5,250 CCTV cameras
Image Credits: Stock Images

On the grounds of the Commissioner’s office in Vepery, Greater Chennai Police Commissioner Shankar Jiwal on Friday officially opened an integrated command and control center (ICCC) as part of the Chennai Safe City Project (CSCP).

In accordance with the plans, 1,750 important locations inside the Greater Chennai Police (GCP) boundaries will have a total of 5,250 CCTV cameras installed, and ICCC will monitor the live feeds from all of these cameras.

The first phase of the project would involve viewing the footage from 4,008 CCTV cameras placed at 1,336 different sites at the ICCC. Furthermore, the Commissioner’s office in Vepery has continuous access to the CCTV camera feeds. Personnel at the ICCC will keep an eye on this.

Read More: Vimeo Introduces AI-Powered Script Generator And Text-Based Video Editor

The Commissioner claimed that a cutting-edge software system powered by artificial intelligence had been installed in the network with several analytics features mixed with alarms to enable quick response to a crime. This technology makes video analytics on crimes including kidnapping, vandalism, vehicle theft, harassment of women, and harassment of women.

According to the Commissioner, the AI-based software will aid in protecting the safety of women because the camera feeds will analyze the scenario using AI software, and the programme will alert the ICCC to take the necessary action even in the case of SOS motions made on the camera.

The video feeds will be replicated and stored in the data recovery center and data center, respectively. The ICCC will be responsible for live stream monitoring, and the same will be made available in the offices of the 12 Deputy Commissioners of Police and the six Joint Commissioners of Police.

Advertisement