Deep Learning Market - Growth, Trends, Forecasts (2022 - 2027)

Deep Learning Market - Growth, Trends, Forecasts (2022 - 2027)

The Deep Learning Market is expected to register a CAGR of 42.56% over the forecast period from 2020 to 2025. Deep learning, a subfield of machine learning (ML), has led to breakthroughs in several artificial intelligence tasks, including speech recognition and image recognition. Furthermore, the ability to automate predictive analytics is leading to the hype for ML. Factors, such as enhanced support in product development and improvement, process optimization and functional workflows, and sales optimization, among others, have been driving enterprises across industries to invest in deep learning applications. Furthermore, the latest machine learning approaches have significantly improved the accuracy of models, and new classes of neural networks have been developed for applications, like image classification and text translation.

Key Highlights

• For instance, consider the new release of PyTorch, Facebook's open-source machine learning framework. Released in October 2019, the new framework, PyTorch 1.3, includes some impressive open-source projects for deep learning researchers and developers. The other new features include experimental support for mobile device deployment, eager mode quantization at 8-bit integer, and the ability to name tensors. One of the noted developments is CRYPTEN, a new community-based research platform designed to address a significant factor that challenges the users from deploying deep learning and machine learning platforms, i.e., security.
• Several developments are now advancing deep learning. According to SAS, improvements in algorithms have boosted the performance of deep learning methods. The increasing amount of data volumes have been supportive of the building of neural networks with several deep layers, including streaming data from the internet of things (IoT) and textual data from social media and physicians' notes. A significant amount of computational power is essential to solve deep learning problems, considering the iterative nature of deep learning algorithms—their complexity increases as the number of layers increases. The hardware running deep learning algorithms also need to support the large volumes of data required to train the networks.
• Computational advances of graphic processing units (GPUs) and distributed cloud computing have put incredible computing power at the users' disposal. This development is led by hardware providers, such as NVIDIA, Intel, and AMD, among others, which have been improving the computational speeds among other features and making them compatible with most-used open-source platforms, such as Tensorflow, Cognitive Toolkit (Microsoft), Chainer, Caffe, and PyTorch, among others. Therefore, 'open-sourcing deep learning capabilities' have become increasingly popular across enterprises. These open-source frameworks enable users to build machine learning models efficiently and quickly.
• In May 2020, NEUCHIPS Corp., an Artificial Intelligence computing company engaging in domain-specific accelerator solutions, launched the world's first deep learning recommendation engine - RecAccelTM - that can perform 500,000 inferences per second. Running open-source PyTorch DLRM, RecAccelTM outperforms inference GPU and server-class CPU by 65X and 28X, respectively. It is equipped with high-bandwidth memory, and an ultra-high-capacity subsystem for embedding table lookup and a massively parallel compute FPGA for neural network inference. RecAccelTM is ready for data center adaptation via a PCIe Gen3 host interface,
• In July 2020, Tencent AI Lab and a group of Chinese public health scientists unveiled a deep learning-based model that could predict the risk of COVID-19 patients developing the critical illness. The procedure was published in Nature Communications. It revised the method in which the lab used the model based on a cohort of 1,590 patients from 575 medical centers in China, with further validation from 1,393 patients. Other tech giants have been taking up similar projects to contain the deadly virus. Using deep learning, Alibaba built a tool for institutions to predict the spread of COVID-19, with almost 90% accuracy rate. Viral structural analysis is performed by Baidu's open-sourced algorithm, which claims the process is 120 times faster than the traditional method.

Key Market TrendsRetail is Expected to Hold Significant Share

• The retail industry has seen a drastic shift in its base of operations in recent times, with many notable brands choosing to reduce the number of onsite offerings in favor of online service. For retailers to remain viable, they need to meet customer expectations, act accordingly, or risk losing loyalty. It is also becoming vital for retailers to adopt burgeoning technologies to make this a reality. Deep learning allows retailers to automate customer experience and streamline processes in a way hitherto unknown. For example, shelf analytics in online scenarios can help with useful recommendations of merchandise and quick classification, which allows customers to make correct choices, with more support, quicker.
• Online retailers such as Walmart are starting to use AI to get product recommendations from customers, but are just barely utilizing the full potential the technology can offer. By using deep learning, retailers can truly harness the power of AI optimizing user experiences and automating time-consuming tasks. For instance, online retailers can use Deep Learning to automatically tag visual data to improve many facets of the user experience. They can use AI to refine the search and return better results to search queries or enhance product images' quality, especially low-quality product photos using color enhancement. Moving forward, retailers can quickly gather data and analyze information automatically using Deep Learning technology.
• A study by Snowflake Computing Harvard Business Review points out that retailers who choose to make data-driven decisions have survived longer. Undoubtedly, retail is rapidly becoming extremely data-oriented. As per the same study, 89% of retailers consider gaining improved insights into customer expectations a significant goal. The models that Deep learning in retail utilizes are sophisticated and advanced enough to handle the challenges that machine learning models fail at. For example, deep learning in retail application models intelligent enough to understand that the release of smartphones with larger screens can eat up tablets' sales. In the case of missing data, deep learning in retail could learn from patterns whether an item isn't selling or is out of stock.
• In January 2020, Johnson Controls announced that its retail solutions portfolio, Sensormatic Solutions, and Intel Corporation, made a collaboration to deliver scalable, AI-powered solutions for retailers. Moving forward, the Sensormatic Solutions AI portfolio at the edge will be based on Intel platforms. Sensormatic Solutions will also leverage Intel Distribution of OpenVINO toolkit and Intel models for delivering its solutions. The AI Vision Intelligence will show drive targeted behavior to improve store operations and shopper experience. It encompasses image processing, deep learning models, and the AI camera developed with Intel to assess associates' responsiveness to customers and measure the method in which customers interact with merchandise and more.

North America is Expected to Hold Major Share

• North America is expected to have a significant share in the global deep learning market, owing to the sustained rise in considerable data volume, coupled with the anticipated increase in the demand for the integration of DL in consumer-centric solutions of the enterprises. The growing emphasis on predicting the key trends and insights related to customer behavior and operations has been a critical driver for significant enterprises to veer toward the use of A.I. and big data for driving value and offering a personalized experience. For instance, Netflix built a machine learning platform based on JVM languages, like Scala. The platform helps break viewers' preconceived notions and find shows that they might not have initially chosen.
• In April 2020, the United States Department of Energy (DOE) announced a plan to provide up to USD 30 million for advanced research in machine learning (ML) and artificial intelligence (A.I.) for both the management of complex systems and scientific investigation. The initiative encompasses two separate topic areas. One topic is focused on ML and A.I.'s development for predictive modeling and simulation focused on research across the physical sciences. ML and A.I. are thought to offer promising new alternatives to traditional programming methods for computer modeling and simulation. A second topic is focused on essential ML and A.I. research for decision support in managing complex systems.
• The United States Department of Transportation formed a new safety regulation to help eliminate blind zones behind vehicles and view people present behind the vehicle. According to National Highway Traffic Safety Administration stats, around 292 fatalities and 18,000 injuries occur due to back-over crashes involving all vehicles. Such regulations are anticipated to encourage the adoption of ADAS, thereby offering opportunities for the region's deep learning market. Furthermore, the region is also seeing an increase in investments from the automakers to develop advanced solutions, driving the growth of the market.
• In February 2020, Micron Technology, Inc., together with technology company Continental, announced plans to enter into a partnership to explore and adapt Micron's deep learning accelerator for next-generation machine learning automotive applications. Automobile infotainment, advanced driver-assistance systems (ADAS), communications, and powertrain control systems are becoming increasingly sophisticated. Micron and Continental will work together in the development of an application-specific version of Micron's deep learning accelerator (DLA) technology designed to be scalable and flexible, at the same time delivering the high performance and low power needed to support industry-standard programming models.

Competitive Landscape

The deep learning market consists of several large players, such as IBM, Google, and Microsoft, among others with substantial industrial experience in big data/analytical platforms. Other new entrants also have been making their way into the market and have been successfully increasing the number of use cases of deep learning across industries. Prominent new entrants that have made a significant impact on the market include H2O.ai, KNIME, and Dataiku.

• June 2020 - Facebook AI Research launched TransCoder, a system that utilizes unsupervised deep-learning in the conversion of the code from one programming language to another. TransCoder was trained on more than 2.8 million open-source projects and outperformed existing code translation systems that use rule-based methods.
• May 2020 - IBM announced that it would apply a range of artificial intelligence (AI) technologies in the automation of the management of IT operations and modernize applications, also known as AIOps. It utilizes machine and deep learning algorithms to time series data, semi-structured logs, structured data, and unstructured data spanning IT incidents and human conversations to track the timeline of an issue.

Additional Benefits:

• The market estimate (ME) sheet in Excel format
• 3 months of analyst support

Companies Mentioned

Facebook Inc.
Google
Amazon Web Services Inc
SAS Institute Inc
Microsoft Corporation
IBM Corp
Advanced Micro Devices Inc
Intel Corp
NVIDIA Corp
Rapidminer Inc

Please note: This publisher does offer titles that are created upon receipt of order. If you are purchasing a PDF Email Delivery option above, the report will take approximately 2 business days to prepare and deliver.