The Benefits of AI Image Custom Model Training

Training custom machine learning models have become a crucial and powerful aspect of processing, understanding, and monetizing visual data today. Based on Artificial Intelligence (AI), this approach allows the fast and effective categorization of massive amounts of images and videos according to the particular needs of a business. 

In contrast to traditional image classification, custom models provide an important layer of flexibility and adapt to the specifics of your company and your industry. The newly trained models operate based on categories that you set — and can thus analyze and organize your image and video database in the best possible way for your particular case. 

In a nutshell, with custom ai model training, your system is taught to recognize concepts from your visual data — the concepts you care about and that hold potential for your business development. You can set any type of category for the classification process, as long as the categories are clear and don’t overlap with each other. This makes you the architect of your visual data classification.

In the sections below, we’ll go over the basics about machine learning and image categorization, and we’ll delve into how custom ai model training based on machine learning actually works. Read on for the full details. 

What Is AI Machine Learning?

Machine learning is a section of computer science and AI that has been gaining exponential popularity. The main focus is the employment of algorithms through which technology mimics the manner in which human beings learn. So, it means that technology can learn and get better at its job with time. Platforms based on machine learning have the capacity to expand their capabilities and knowledge in unprecedented ways — with precision and accuracy growing with every bit of new data processed. 

In layman's terms, the machine learning term entails ‘learning by doing.’ In technical terms, machine learning algorithms are taught to classify data and make predictions based on statistical methods. This makes them a powerful tool for unraveling insights from data that would take years to process otherwise. Equipped with these in-depth insights, business leaders and managers can make well-informed decisions that drive business growth and development. 

Some popular uses of machine learning that we are already experiencing — even in daily life — include recommendation engines (like the series and film suggestions you get from your streaming platform) and self-driving motor vehicles. 

1. Machine Learning and Deep Learning

Going a step further in understanding ai machine learning, it’s good to present the term ‘deep learning’, too. Often, machine learning and deep learning are seen as identical, but in fact, deep learning is a sub-field of machine learning. It entails a different learning process and has been referred to as ‘scalable machine learning’ by Lex Fridman

The main difference between the two is that deep learning doesn’t require labeled datasets to learn. It can process data in raw and unstructured form, too. This makes it more independent from human input — and able to process larger amounts of data. Traditional machine learning, on the other hand, needs more actions from humans, and, in particular, more structured sets of data for learning. 

What Is AI Image Categorization?

Image categorization, sometimes also referred to as image classification, is powered by computer vision. It employs machine learning and image processing to sort images and videos by distributing them into categories, which are usually set in advance. It may as well be one of the most significant elements of digital image analysis today.  

Image categorization is widely used in a number of fields. Most notably, it’s the basic tool for automating content moderation online. However, it has numerous other uses, such as database sorting, product discovery in the field of commerce and retail, and asset management in technology and cloud services, among many others. 

In essence, the powerful business use of image categorization is that it allows you to gain control over huge image sets. The engine is taught to discern different categories through a set of local and global visual features. Once it learns them, it’s able to spot the precise category for a new visual that it processes.  

To get an idea of how Imagga’s image categorization engine works, you can check out our Visual Categorization Demo. Our Image Categorization API boasts a couple of powerful features: it’s accurate in its classification; it’s scalable even for enterprises; it’s simple and adaptable for cloud, on-premise, or edge; and it’s customizable to the specific needs of your business. 

How Does AI Image Custom Machine Learning Model Training Work?

Creating custom models is at the heart of performing effective analysis of specific visual data that businesses need today. Pre-set categories don’t always satisfy these needs — and hence, customization becomes key. 

With AI image custom model training, you can specify precisely the categories that your visual content has to be distributed to. The number of categories is limited, theoretically, but in practice, Imagga’s custom training can handle training with tens of thousands of categories.

The custom-trained model, tailored to the specifics of your business, can then be paired with the Image Categorization API, providing you with a powerful tool to classify visual data and maximize its use and impact. 

1. AI Image Custom Machine Learning Model Training Steps

  1. Feeding with datasets: The training datasets have to be inserted into the engine, containing sample visual data for each category. The categories have to be non-overlapping and straightforward. 
  2. AI Model training: Our machine learning experts build a deep learning classification model based on your data and the specific categories that you set. Content that has already been classified is used for the training so that the engine gets accurate principles for categorization. Afterwards, it can process new content and automatically classify it according to the newly created categories. 
  3. Deployment in your systems: The newly trained model is embedded in an API that is seamlessly integrated with the systems and workflows that your business is already using. Through a customer demo, you can evaluate the model in advance. 

PlantSnap Case Study: Image Categorization and Custom Training at Work

A great example of the powerful combination between Imagga’s Image Categorization API and Custom Training is the case study of our work for PlantSnap.

PlantSnap is an app that helps people identify any plant anywhere on the planet. It’s an amazing knowledge base where you can find all kinds of flora species. You just take a photo of the plant you want to identify — and the app provides you with its name and information about it. 

To power up the plant recognition, PlantSnap needed an image categorizer that could handle a massive amount of categories — as there are 320,000 different species worldwide. Most image recognition providers couldn’t address this need, as they couldn’t train such a huge amount of categories, and couldn’t guarantee accuracy decreasing due to the large volume. 

At Imagga, we decided we were up to the challenge. We invested in getting the DGX Station from NVIDIA, powerful hardware that we paired with our outstanding computer vision technology. The result is all that we expected it to be. 

The custom model that we built for PlantSnap is ten times faster in training and doesn’t compromise accuracy. It’s combined with our state-of-the-art image categorization API that boosts high accuracy rates. 

In the deployment process, we successfully resolved another challenge: plant look-alikes. Even plants with similar visual characteristics can now be identified and discerned by Imagga’s custom-trained model. 

As a result of our efforts, Imagga is now a core technology in the PlantSnap app — boosting accurate plant recognition for all types of flora species worldwide: 320,000 plant classifications with a 90% precision rate for the top 5 results of each search. 

Get Started with Your Image Custom AI Model Training 

Due to our extensive experience in providing image recognition and categorization tools for companies from a wide variety of industries, we’re equipped to create powerful, tailor-made solutions.

Want to start building your custom-trained model for image classification? Get in touch to learn more about our solutions from Imagga’s experts. 

Frequently Asked Questions

How do you custom-train image AI models?

Custom training of image AI models relies on machine learning to train the engine to discern visual data from data assets and classify it into custom categories. 

The process involves three steps. The first one includes providing existing datasets and concisely formulated, non-overlapping categories for their classification. Next, deep learning is used to build a classification model with the custom data, based on existing classified content. Then, the model is plugged into an easily embeddable API — which can start processing new visual data. The more images are analyzed and classified, the better the engine becomes at categorization. 

What are the benefits of AI Image Custom Training?

Image custom training based on AI brings unseen advantages to businesses from a number of venues. Companies that operate with huge amounts of visual data, such as user-generated images and videos or crawled visual content, need effective ways to sort and arrange the data. Manual categorization and processing is unthinkable and overwhelming due to the massive quantities. 

This makes machine learning algorithms indispensable help in this process. Image custom training, in particular, offers tailor-made solutions for the specific needs of a company. Instead of using pre-set categories for image classification, it allows a business to take control of its visual datasets by providing it with a deep-learning classification model that is trained to work with its specific categories and data.  The deployment of AI image custom training is robust and can work with any dataset size, while also being flexible and adaptable to cloud or on-premise solutions. 


Imagga Technologies Awarded Prestigious €2.1M Grant Funding for AI-Driven Content Moderation Platform Development

Imagga Technologies is excited to announce its receipt of a pivotal grant under the "Support for Innovative Enterprises awarded with 'Seal of Excellence'" program, a constituent of Bulgaria’s Recovery and Resilience Plan. This esteemed endorsement empowers Imagga, a leader in image recognition and artificial intelligence solutions, to forge ahead with its innovative AI-Mode project.

The AI-Mode project aims to create a state-of-the-art Content Moderation Platform, designed to revolutionize how companies approach and automate their content moderation workflows. By harnessing advanced AI capabilities, the platform will facilitate proficient image, video, and live streaming moderation, providing an indispensable tool in maintaining the sanctity of user-generated content (UGC) platforms and safeguarding brand integrity.

The Imagga Content Moderation Platform is a full-circle content moderation solution combining the best of artificial and human intelligence. This robust system is proficient in identifying and filtering a broad spectrum of inappropriate content, including explicit imagery, offensive symbols, regulated substances, violence, and propaganda, thereby reinforcing a brand’s digital ecosystem. The innovative use of machine learning significantly amplifies efficiency, potentially reducing the volume of concerning content by a factor of twenty through proactive AI-assisted filtering and iterative refinement from human moderators.

This development project originates from Imagga’s successful application and subsequent distinction under the European Innovation Council (EIC) Accelerator program, which has been operational since 2018. The EIC Accelerator is a cornerstone of the European Union's commitment to fostering innovation, offering cutting-edge enterprises the chance to compete for up to €2.5 million in grant financing, complemented by an opportunity for an equity investment of up to €15 million. Despite the rigorous competition with over 75,000 applicants, Imagga has emerged as a standard-bearer of innovation.

The Seal of Excellence is awarded to project proposals submitted under a Horizon Europe call for proposals and ranked above predefined quality thresholds but not funded by Horizon Europe due to budgetary constraints.

The Seal of Excellence is a quality label first introduced during Horizon 2020, the EU’s research and innovation framework program (2014-2020). It has gradually become a key instrument in recognizing innovation on EU level, which is later on funded by local government grant schemes. 

So far, only 12 Bulgarian companies have received SoE, and the funding will come as a grant under the fourth measure of Bulgaria’s Recovery and Resilience Plan - "Support for Innovative Enterprises awarded with "Seal of Excellence" and co-financed by NextGenerationEU. It’s worth mentioning all applicant projects are assessed by four independent remote assessors (among 1,500 professionals authorized by the EIC after assessment of relevant expertise). Their assessments are combined in a general report with a maximum integral assessment of up to 15 points. Companies with a score of more than 13 points are awarded the 'Seal of Excellence'.

Fuelled by the financial stimulus of the AI-Mode grant, Imagga is committed to developing a sophisticated Content Moderation Platform that aligns with stringent EU content regulation standards and pioneers real-time moderation capabilities. This development underscores Imagga's commitment to fostering a safer, more respectful digital space, fortifying its position at the forefront of AI-driven technological solutions.

This publication was created with the financial support of the European Union - NextGenerationEU. All responsibility for the document’s content rests with Imagga Technologies OOD. Under no circumstances can it be assumed that this document reflects the official opinion of the European Union and the Bulgarian Ministry of Innovation and Growth.


CounteR Technical Meeting in Sofia

At Imagga, we were proud to co-host CounteR's fifth technical meeting in Sofia alongside the European Institute Foundation. For three days (September 26-28, 2023), all consortium partners reviewed the progress of the CounteR Project and further tested and validated the software tool in collaboration with our law enforcement partners. We took a deep dive into evaluating our past and current efforts and strategized for the upcoming tasks within our various work packages (WPs).

Our kick-off day was particularly illuminating, with an insightful review of WP2 – "Social and Psychological Factors in the Radicalisation Process.” We noticed an interesting paradigm shift in the research approach towards preventing radicalization. Instead of solely focusing on individual radicalized identities, the emphasis is gradually leaning towards understanding the underlying issues of radicalization hubs. Our discussions extended to WPs 3 to 6, where we explored enhancements in data collectors, image processing, social network analysis, and the more technical facets like semantic reasoning, deep reinforcement learning, and network algorithms for illicit content removal.

Day Two saw us delving into the intricacies of WPs 7 to 9, emphasizing data privacy and ethical standards. We reflected on the insights gained from our first pilot in Milan in June 2023 and used those learnings to refine our approach for the imminent second piloting session.

The culmination, Day Three, was primarily hands-on. Through various test scenarios and datasets, the tool's efficacy was assessed. A successful stress test was performed after the law enforcement representatives uploaded the synthetic datasets.

A significant portion of the testing was centered on the NLP analysis module, an excellent work by INRIA. Collaborative discussions between Insikt and us at Imagga enriched the process, especially concerning report issuance for image analysis. This was pivotal in enhancing the system's subsequent versions. Insikt and Imagga discussed the appropriate ways to issue reports for image analysis to have better and richer user feedback.

The culmination of these sessions wasn't just progress checks. We robustly tested our system alongside our law enforcement agency partners, gearing up for its grand release.

Furthermore, we dedicated a reasonable amount of time in Sofia to deliberate on CounteR's communication strategies and commercialization prospects. By the project's 36th month, we aim to launch a comprehensive business strategy to promote the CounteR solution. The knowledge amassed from CounteR isn't just technological; it spans across social, legal, and policy sectors – all of which are pivotal for our future roadmap.


A Day of Innovation and Insights: Imagga at the AI4Media Workshop in Amsterdam

Amsterdam was abuzz with action and innovation, and we at Imagga were right at the heart of it!

As part of the transformative AI4Media project, the Speculative Design Workshop was convened, and Georgi Konstantinov, CTO of Imagga,  had the honor of representing us at this prestigious event. Our mission? To showcase Imagga’s cutting-edge solutions for content moderation and organization. We explored current uses of our technology and ventured into the tantalizing possibilities of what the future might hold.

But the real magic of the workshop was how it became a melting pot of analytical reasoning and out-of-the-box creativity. A lot of this was thanks to the unparalleled skills of our facilitator, Nina Pavlovska from ZEZA. Under her guidance, a cascade of innovative ideas emerged, promising to shape the direction of the AI4Media project in ways we had only dreamed of.

But the excitement didn’t end there! As fortune would have it, the workshop’s timing overlapped with the Mozilla Festival Amsterdam. This serendipitous occurrence allowed us a golden opportunity to test and validate our freshly minted concepts with a broader audience.

In a surprising twist to an already exhilarating day, Georgi stepped onto the stage for ‘The People Speak’ – a live talk show. Representing our work at Kelvin Health, he shared insights into AI’s pivotal role in healthcare. The conversation was both enlightening and humbling, emphasizing AI’s vast potential and responsibility in shaping our healthcare systems.

Reflecting on the day was a whirlwind, a marathon of thoughts, ideas, and interactions. But it also underscored the limitless horizons that lie ahead for AI, Imagga, and the larger digital community. Here’s to many more such days of discovery and growth!


The Evolution, Importance, and Challenges of Automated Media Organisation - White Paper

The modern digital age has ushered in revolutionary advancements in Machine Learning (ML) that have reshaped the media landscape. These advancements have significantly fueled the potential of automating extensive media collections, promising efficiency and precision like never before. A mere decade ago, the ambit of image recognition was limited to basic shapes and objects. Contrast that with today, where our image recognition AI allows instant recognition of a plethora of visual content.

However, despite these impressive strides, the journey of AI and ML in media isn’t without its stumbling blocks. Imagga’s eye-opening survey indicated that only 7% of media companies leverage automated solutions for visual content organisation. The majority still rely on the age-old manual cataloging methods, essentially anchoring themselves to human limitations.

Addressing Industrial Needs and Overcoming Technological Hurdles

Every media entity, regardless of size, grapples with the daunting task of efficiently organising the sheer volume of visual content it generates or receives. This content could range from in-house photo and video collections to user-generated content (UGC), which has exploded in popularity and scale in recent years.

Traditionally, media houses have resorted to manual tagging and categorization. However, such practices have proven to be sub-optimal, especially in a world where speed and accuracy are paramount. A paradigm shift to automated AI-powered tools that promise optimal organisation is obviously needed for media companies to face the challenges and adapt to new realities. 

Yet, the path to automation is lined with challenges. Introducing such advanced tools demands hefty costs, a deep pool of AI expertise, and the crucial need to manage expectations. It’s vital for the non-technical staff in media companies to possess a grounded understanding of AI’s capabilities and limitations. Unrealistic expectations can hinder progress and inflate investment without yielding the desired results.

Unlocking Value from Existing Content

Amidst the clamor for the new, there’s a treasure trove lying in the old. Media entities sit on vast reserves of content that, with the right tools, can be repurposed, opening up new avenues of value and revenue. However, the real challenge lies in making this content easily discoverable.

Imagine a scenario where content that was created for one specific use-case can be easily accessed, modified, and reused for another. This ensures cost efficiency and maximizes the return on content investment. The magic key that unlocks this potential? Proper metadata and archiving practices that are bolstered by AI.

Looking Forward: Adapting to the New Era

Much like any other, the media industry isn’t isolated from disruption. With the exponential rise of user-generated content and citizen journalism, traditional revenue streams are being constantly challenged. This disruption has made it imperative for media companies to adapt and realign their strategies.

Media entities must rethink their content creation models in a world where every individual with a high-quality camera phone can be a content creator. This is where AI-enhanced tools step in, serving as the bridge to the future and helping media entities remain competitive, relevant, and efficient.

The Path Ahead: AI for Media

The marriage of AI and media promises a future brimming with potential. But to harness this potential to its fullest, there’s a need for a new, AI-based approach. Media entities must embrace technological upgrades, invest in staff training, and constantly re-evaluate and evolve their existing workflows.

As we stand on the cusp of this transformation, it’s essential for AI developers and media companies to foster a spirit of collaboration. This mutual partnership will ensure that technological capabilities align perfectly with the industry’s ever-evolving needs, ushering in a new era of efficient, innovative, and forward-thinking media.

Download the latest Imagga White Paper on AI Technology for Image and Video Organization

Imagga’s white paper is sponsored by AI4Media, a Horizon 2020 project, building an AI network for organisations from academia and industry that embraces a vibrant ecosystem focused on AI for media and society and enables the quick market uptake of technologies and research/business collaboration opportunities. A series of white papers are published in an effort to align AI research with the industrial needs of media companies. These white papers describe the most critical challenges and requirements for AI uptake in different media sectors. 


Imagga Attended CounteR’s Autumn Technical Meeting in Malta

Imagga took part in CounteR Project’s third technical meeting in Valetta/Malta on September 28-29, 2022. The meeting brings together representatives of the consortium partners of the CounteR Project. 

Georgi Kostadinov, CTO of Imagga, together with representatives of ELTE University and Insikt Intelligence, presented the progress on WP4/Data Analytics for Detecting Radical Content‘s data understanding and preparation, the basic features and transfer learning regarding the NLP analysis, and image and social networks analyses.

The partners are reviewing the status of the work with a focus on current tasks. The participants are also planning ahead and discussing the upcoming project deliverables to be submitted by Month 18.

The event is a perfect opportunity for all consortium partners to meet, update each other on the project's progress, exchange ideas, get inspired, and create business opportunities. 

The CounteR consortium brings together an illustrious group of international subject-matter experts in counterterrorism, radicalisation, and privacy law, six European law enforcement agencies with practical in-field knowledge, as well as a group of technical SMEs and academic partners.


Imagga attended the IMPROVATE Defence and Cyber Conference in Sofia

Georgi Kostadinov, CTO of Imagga, along with Borislav Mavrov, Programme Director at Europеan Institute, took part in the IMPROVATE conference, a high-level event held in Sofia, Bulgaria, with the attendance of AI experts in the fields of defense and cybersecurity, as well as leading politicians, government, industry, CSO, academia, and diplomatic community representatives from NATO, Bulgaria, North Macedonia, Israel, Greece, Bosnia, Montenegro, Romania, and Albania.

“Always happy to discuss the potential of Imagga's pioneering computer vision tech in countering radicalisation and eliminating online threats developed as part of the CounteR Project”, commented Georgi Kostadinov, CTO of Imagga. 

Excellent sessions from leaders such as Garry Kasparov, Rosen PlevnelievDr. Solomon Passy, and many AI experts in defense and cybersecurity!

IMPROVATE is a platform that makes technology and innovation accessible to new markets around the world. Launched in September 2020, IMPROVATE operates out of London and Tel Aviv and serves as a platform to connect leaders, decision-makers, companies, and investors with technology and innovation. IMPROVATE’s Board members and partners are: former world chess champion Garry Kasparov, President Rosen Plevneliev – former President of Bulgaria, Prof. Zeev Rotstein, - Medical systems expert, IDF Major Gen. (Ret.) Amos Gilead, Yigal Unna -Director General of the Israel National Cyber Directorate (2018-2022). Victor Ponta - Former Prime Minister of Romania | Yves Leterme - Former Prime Minister of Belgium | Dr. Vlado Buckkovski – Former Prime Minister of North Macedonia.


future of content moderation

Content Moderation - Future Trends & Challenges

During the last few years, content moderation powered by Artificial Intelligence has grown exponentially. It has developed in ways that were unimaginable just a decade ago, breaking concepts and widening the horizons of what’s possible. In this article we'll have a look at the future of content moderation.

future of content moderation

Automated content moderation has been fueled by ever-evolving machine-learning algorithms that constantly improve in accuracy and speed. Just 10 years ago, image recognition was only able to classify and detect basic objects and shapes. Now, thanks to the advancements of deep learning, image recognition algorithms for instant detection of all types of inappropriate visual content are a reality.

Automated (also referred to as semi-automated) content moderation thus offers important new capabilities for businesses of different venues that need effective screening of digital content. The AI moderation platforms address a number of key challenges that online platforms and companies face, including: 

  • Huge amounts of user-generated content need to go online immediately, but still have to be monitored for appropriateness, safety, and legality. This can make it difficult for online platforms to grow and scale internationally if they don’t have an effective way to screen all postings — textual, visual, and even live streaming. Without moderation, these businesses risk great reputational harm, along with a list of other negative consequences. 
  • Content moderation has to happen in real-time, which is especially difficult for live streaming and video that are becoming the most popular content formats. The complexity of screening visuals, texts and moving images at the same time is tremendous. 
  • User safety and especially the protection of vulnerable groups is becoming a priority in legislation that covers digital platforms. This means that in many places across the globe, online businesses are required by law to have solid Trust and Safety programs and protection mechanisms based on content moderation. This is necessary not only to ensure the upholding of their internal principles and guidelines, but to safeguard consumers. 
  • The stress and harmful effects on human content moderators from exposure to shocking, violent and disturbing content is significant. Digital businesses aim to minimise these negative consequences and to protect their moderating teams from the worst content. 
  • Digital platforms have to be able not only to scale in terms of countries and amounts of content that goes live, but also to adapt to quickly changing circumstances and norms for content appropriateness. 
  • Public manipulation, political propaganda, disinformation through fake news, and the rise of DeepFakes are disturbing yet prevailing new phenomena online. Both official authorities and online platforms need an effective way to fight them, and machine learning algorithms are the key to that.  
https://youtu.be/kmHqqf2ztJw

Research challenges 

While holding great potential and already showing impressive results, there are challenges that AI-powered content moderation is facing. 

One of the major issues with which automated content moderation is struggling is recognizing context. Machine learning algorithms can find it difficult to differentiate between subtle cultural and social trends and phenomena. For example, if the algorithm is set to remove all nudity, this is what it would do — even if the nudity is related to art or important news pieces. A prominent example was the case from 2016 when Facebook removed the photo of the iconic Vietnamese ‘napalm girl’ who is naked

Another important challenge that AI platforms need to overcome is multilingual moderation. While they are getting better at it and are surely improving how content moderation in different languages is conducted, there are still obstacles on the way. The process is not only about acknowledging the direct meaning of words and phrases, but their social and cultural connotations that may make them offensive and inappropriate. In this respect, the more feedback machine learning algorithms receive, the better they can become at spotting the nuances in content — which is definitely not a mission impossible, but simply a gradual process that takes processing large amounts of data. 

Live streaming and live video are another interesting challenge for AI-based content moderation platforms. They generate such a substantial amount of data per second that manual moderation is simply an impossible task. Moreover, applying AI on each frame of the live broadcast generates high platform costs. A fast and accurate AI needs to be developed to overcome these hurdles to efficient and cost-effective moderation of live streaming and videos.

Societal and Media Industry Drivers

Merry is a fact-checking manager at an online newspaper with a solid and long-standing reputation for trustworthiness. Her job is to ensure every piece of information that gets published through the media’s channels is accurate. She needs to check not only facts and textual references, but also visuals. In a way, Merry is a modern-day fighter against disinformation — the plague of today’s digital world. 

But doing all of this on her own — and even with her fact-checking team — is a monstrous task. The process simply takes enormous amounts of time and effort. That’s why Merry needs a viable and scalable solution to checking accuracy and preventing the spread of fake news and visuals. 

This is especially important during the current national election. Politicians, ordinary citizens, and people who want to spread fake news are all posting about the topic constantly — and it’s very difficult to sift through what’s true and what’s not. In particular, a scandalous story about financial misdealings that one of the party leader’s is involved in is spreading online — mostly fueled by his political opponents. Merry has to figure out how true the lead is and whether to publish it. 

With the help of an AI-powered content moderation platform, Merry can screen various materials around the topic for authenticity. She can catch textual references, as well as photos that have already been posted, for example.  

Daniel is a content editing manager at an online news outlet. He’s in charge of guaranteeing that all published content complies with the standards of the media and the legal framework. His most challenging task is to ensure the compliance of live streamings. Catching inaccurate and harmful video content in real time is a tough nut to track. Without technical support, it is a burdensome task to monitor live streaming content as it occurs. 

This is where a content moderation platform based on machine learning algorithms can kick in. It processes all live video streamings, checking them for inappropriate verbal and visual elements. If there are such, the platform can immediately signal to the editors. 

Daniel can test the capabilities of the AI platform in practice during an important live streaming with a local politician at a public rally. The situation is uncontrollable, as it’s a place full of people where anyone can appear and take over the stage. With the help of the moderation platform, Daniel can have the live streaming screened for problematic content throughout the whole event.

Content Moderation Future Trends and Challenges

Future Trends for the Media Sector 

The role of content moderation in the media sector cannot be overestimated — in fact, it’s crucial for its wholesome development on a couple of levels:

  • As illustrated in the vignette, content moderation algorithms are the key to fighting the widely spread online disinformation. They can spot inaccuracies in textual data and fake or old visuals and videos. These capabilities can be a gamechanger for the media sector that direly needs adequate fact-checking in the oceans of information that get published. 
  • On the basis of the massive volumes of content that machine learning algorithms process, they can also make trend predictions about the types of content that needs to be moderated. This can be of huge importance for combating harmful tendencies in user-generated content sharing. 
  • Content moderation, if done in-depth, can help in detecting the intent of disinformation in order to differentiate between positive (for example, to keep state secrets) and negative (to influence opinions and harm society).

Goals for Next 10 or 20 Years

The long-term vision for AI-powered content moderation is a truly ambitious one. 

First and foremost, content moderation would need to leave the semi-automatic status it currently has. To be fully useful, scalable, and powerful, it should be more autonomous. The vision is that AI-powered content moderation platforms would become a monitor that is always on and oversees any and all content that goes online. It would screen for all types of abnormalities to ensure protection of users from harmful and illegal content and a safe online environment without fake news — thus taking care of everything from violence and nudity to propaganda, radicalization and disinformation, and all that’s in between. 

The second big goal for content moderation’s evolution is self-learning — which is already in motion, but can reach new heights. With the data that is being fed in the moderation platform in real time, the machine learning algorithm becomes better and better. It expands its knowledge base with practical examples and input from moderators. With time, this is how the AI can become more independent from humans in terms of feedback loop. In the foreseeable future, this can reach a point where the moderation platform becomes an autonomous machine that identifies and filters content accurately and effectively with no human input. 

A third long-term goal for content moderation is the creation of instant and efficient on-device moderation. Nowadays, moderation is executed on the server end, only after a piece of content has already gone live. This means that harmful content can sneak in for a moment and be accidentally shown to users. In the near future, moderation would be possible on the customer device itself. This would happen before the content has gone live. This advancement would enable the prevention of illegal and disturbing content appearing on the device level, thus ensuring full protection for end users.


Imagga to be part of CounteR – countering radicalisation for a safer world

We are happy to announce that Imagga is taking part in a consortium working on the critical topics concerning democracy - identifying radical content online and creating an integrated situational awareness system to counter violent terrorism and enable crime prediction. 

The CounteR solution, an open platform for analysis and early alert, will be launched under the project. The platform will collect and analyze data from dispersed sources in order to predict critical communities at risk of radicalization and violent extremism and aid law enforcement to detect radicalization. 

Radicalization is one of the significant challenges online for societies in Europe and beyond. It’s easier than before for people to be misinformed or exposed online to extreme social, religious, or political ideas. Especially vulnerable are some groups with unmet psychological needs for belonging, status, or such with mental illness. The ease of communication and reaching out to multitudes of possible prospects via social media platforms make this threat imminent. Terrorist groups that use social media for radicalization have mastered their tools – it is getting harder to detect such efforts due to the complexity and mass spread of such problematic content.  Early detection and flagging problematic content might be one of the most essential tasks in safeguarding Europe and its democratic societies. A combined effort in adopting anti-radicalization policies, technologies, and tools is becoming super important in the fight to counter violent extremism online.

To address this imminent problem, the CountR consortium will work on a platform that will combine critical technologies and know-how to make it possible for real-time detection of propaganda and radicalization in user-generated content. The various participating partner technologies will achieve the robustness of the detection methods. Imagga will contribute to our advanced image and video image analysis and efficient live-stream Content Moderation, which, combined and cross-referred with text analysis, will make it possible to detect radical messaging even though each media type might not have that particular message. 

Imagga’s contribution to the CounteR project is the development of instruments and deep learning algorithms to analyze images used for the automatic detection of radicalization and extremist content. Imagga’s image recognition expertise and proprietary technology for Content Moderation will be used to build robust mechanisms for detecting problematic content. This includes detecting specific keywords, a scene in an image or an action in a video that might be related to or inclined to radicalization.


A year in pandemic: 2020 Imagga recap

2020 was not the typical business year. With the announcement of COVID-19 pandemic in the beginning of March 2020, our business and the lives and business of many others have been transformed. In a way it restarted in a very wired way our plans, activities and work practice. March 2021 marks one year from the announcement of the pandemic and we felt now is more appropriate time to reflect on what have happened one year surviving the pandemic. 

COVID-19 changed a lot in our personal lives and impacted our company, and our partners. We all had great experiences but the world pandemic, unpredictable and unexpected, stormed into our lives. Historically, with difficult times comes much innovation. The good news is we are built to stay and we adapt and change, learn from the hardship and end evolve our lives, product and services to better match the new reality. 

The following is a recap of Imagga's progress since the turning point in March 2020 when we realized we are operating in the reality of world pandemic.

Kelvin Health Spin-Off

Triggered by the COVID-19 pandemic, we joined forces with outside experts. We started together a spin-off called Kelvin Health that applies Imagga's AI technology to aid the screening and monitoring of medical conditions with the help of thermal imaging. A group of enthusiastic medical doctors proposed to validate their observations that AI-enabled analysis of thermal images of the chest and back can help better understand the development of COVID-19 and its complications. We signed agreements for data collection with several prestigious hospitals in Europe. This is helping us to significantly improve the analytical part of the solution enabled by Imagga's machine learning tech. 

The scope of the project went far beyond just COVID-19 and we are currently working on helping monitor and screen various cases of breast cancer. Soon to be announced partnership with a pharma company that is at the forefront of fighting cancer worldwide will help us gather valuable data to aid the AI analysis.

 

Core Technology Advancement  

During last year one of our priorities was Imagga's Content Management Platform. We released a private beta version of the CM platform that offers automatic filtering of unsafe content in images, videos, and live streaming. We've put significant effort into improving the dataset quality, increasing accuracy. Thanks to pilots with several significant customers, we managed to improve the platform's stability and extend it to cater to the exponentially growing demand for live streaming moderation. 

Our software stack and deployment strategies got a significant upgrade last year. We've improved the overall performance, including a better migration process, optimized and more stable software stack. This resulted in even better performance on our Cloud API and on-premise solution, increasing their stability. 

EU Support for Technology Advancement 

2020 put an end to a fascinating project, funded by the EU: Photobox - personal photo organization enhanced by Imagga's image recognition and image classification technologies. We've managed to significantly improve our image datasets and make the organization of personal photos more precise and accurate. We will be integrating Photobox technology into our flagship products, the Tagging and Facial Recognition API’s, to make them even more powerful and exciting for our variety of customers. 

We've started a collaboration with ZAZ Ventures, leading deep-tech innovation grant consultants. Creating an excellent grant proposal sometimes goes beyond the workforce abilities of technology startups. It often requires tapping into an elaborate grant proposal framework and/or participating in a consortium of business partners working on a multi-industry solution. 

We've got the Horizon2020 Seal of Excellence. Imagga is one of the two companies in Bulgaria for the past two years being recognized with that recognition. 

2020 also brought two new EU Commission-funded projects to us. AI4Media is a high-impact project run by a consortium of 30 renowned academia, research, and enterprise organizations. AI4Media, which is funded with a total budget of 12M EUR by the European Union's Horizon 2020 Research and Innovation Programme, will focus on delivering the next generation of core AI advances to help media meet the constantly growing requirements but also comply with the core EU values of trustworthiness and ethical engagement.

The second successfully funded project that we are part of is CounteR. CounteR brings data from dispersed sources into analysis to create an early alert platform for data mining and prediction of and support the fight against radicalization and prevent future terrorist attacks in the European Union. Project partners are several police agencies and NGOs and academies doing in-depth research on public violence and terrorism prevention. 

 

Customer Integrations

Somebody once said there's no better investment than a paying customer. Nothing makes us happier at Imagga but to see the needs of our customers met, enabling them to automate "things that are impossible to do manually." 

One of the most significant integration achievements of 2020 is the plant lense of the Plantsnap app right into the interface of Snapchat. That was made possible thanks to Imagga's very high-performing deep learning cloud API meeting Snapchat's demands. 

Eden Photos is another success story of 2020 - bringing the power of image tagging to Mac and Windows desktops. Imagga's tagging and categorization APIs enable Eden Photos to provide automated tagging and organization capabilities to your personal photo collection. 

Imagga processed 600M images in the last 12 months via it's cloud API and even more on-premise in helping our customers improve their end-user experience, monetize their visual content, or assist them in building custom solutions based on visual content. 

Last but not least, having a strong team is the best asset every company can have, especially in challenging years such as 2020. Team of people who are ready to fight adversity to support our common mission and keep conquering new horizons together in the long run. Efforts of the team did not go unnoticed. Georgi Kostadinov, our Head of AI, was featured in Forbes “30 under 30” in Bulgaria for 2020.

2020 was challenging but at the same time very exciting for Imagga, and we believe for many businesses worldwide. It brought both the stress and the freedom coming from uncertainty and marked the efforts we need to put our attention in - fighting for a better, more secure future, and Imagga's technologies are a big part of it. We are excited about how 2021 is already unrolling! The best is yet to come!