Shared todayXR Industry SeriesIn-depth Research Report:XR Industry Report: 3D Content Breaks the Hardware Linkage, and the XR Industry Chain Welcomes New Development Opportunities
Report Producer: Bank of China**).
Report total: 34 pages.
Featured Report**: The School of Artificial Intelligence
3D content: Digital content with in-depth information is the upgrade direction of information carriers, and the Xinhui tailor is developing in the direction of efficiency and growth
The development of human information communication carriers has gone through four stages: sound, text, and **. Sounds and text are ID information, and ** are 2D information. The law of information development is that the density of information is getting bigger and bigger, and the way of information presentation is becoming more and more intuitive, which can speed up work efficiency and enhance the intuitive sense of communication and entertainment, so it has become the direction of information carrier iteration. We believe that the next stage of information carriers will be able to portray the real world more vividly, which can further enhance work efficiency and increase communication speed, and 3D content may become the general trend.
3D content refers to content with vertical dimensional information, which can produce three-dimensional perception compared to flat 2D content, 3D content has vertical dimensional information, which allows people to produce stereoscopic visual effects. A typical example is a 3D model, volume**, or 3D game. Generally speaking, 3D content is stored on a computer, and content producers can improve content production efficiency by reusing 3D content. Typical 3D content production methods include volumetric shooting, 3D game production, etc., the latter is produced through industrial processes such as modeling, texture, animation, etc., which requires a professional team and a complex process. Another type of content that achieves 3D perception is actually 25D content. 2.5D content itself does not have depth information, but it can be combined in a clever form of information to produce a 3D-like effect. Typical 25D content includes panoramic D movies. Taking Panorama as an example, the principle is to splice multiple ** together and project them onto an equidistant cylinder or sphere, allowing users to achieve a 360"** effect. The principle of 3D movies is to allow the user's left and right eyes to receive slightly different 2D ** binocular parallax through polarization, etc., to form stereoscopic vision. 2.The production method of 5D content is mainly multi-camera shooting, and the content production is achieved through algorithm fusion processing.
3D content has the effect of improving efficiency and speed, perceiving overthrow, etc., which is the upgrade direction of Xinhui's tailoring
3D content provides more realistic and vivid visual impact, helping to work faster and faster. (1) It can provide more intuitive visual effects. By rotating, zooming in, and other operations, users can observe the products displayed in 3D from different angles, obtain more comprehensive information, and achieve results that are not possible with conventional 2D content. (2) Support spatial interaction and communication. 3D content contains depth information and is generally accompanied by a spatialized interface, allowing users to interact with 3D content more naturally. (3) It can improve work efficiency and accuracy. The greater visualization of 3D content reduces communication errors, effectively conveys work information, and maximizes time and money savings, thereby increasing efficiency.
XR equipment is the best hardware carrier for 3D content, and the production difficulty leads to a small amount of content
3D content display lacks hardware conditions, and XR is the best display medium for 3D content. Usually the physical screen that the user touches is two-dimensional. Depth information cannot be displayed. There are currently fewer technical forms that support 3D content, and XR devices are one of the main ways to support 3D content. XR uses two screens to create a binocular parallax similar to the real world, which can better reproduce the sense of 3D space. With XR devices, users can walk, interact (VR), or overlay 3D content in real-world space (AR) for a greatly enhanced user experience.
The low amount of 3D content is a big reason for the slow growth of the XR industry. There are two main ways to produce content for XR: scene construction based on 3D geometry and image-based scene construction. The former is a specialized, highly technology-intensive task, while the latter often requires specialized equipment to produce, both of which limit the amount of 3D content available. As of 2022, the total number of content on major VR platforms is only more than 10,000, mainly due to the difficulty and high cost of 3D content production.
Since the release of Vision Pro, industry chain companies have launched a series of software and hardware upgrades to accelerate the improvement of the 3D content ecosystem.
Apple introduced Spaces** and supported the iPhone as a tool for creating 3D content. During WWDC2023, Apple unveiled Space**, a new form of 3D content in Apple Vision Pro that can record and Spatially. The audience feedback of the space ** was good, and the following comments were given by reporters such as "Wall Street**": "Space** could become the killer app of Vision Pro", "Realistic enough to make the hairs stand on end, and it can create intimacy in a two-dimensional**and**unattainable way", "Space** is more like a memory than **", and the effect exceeded the expectations of the experiencer.
But for the masses, shooting with Vision Pro is not convenient. Apple announced at its Fall 2023 launch event that the iPhone Pro series of smartphones will support Recording Space**, and iOS 17 in early DecemberVersion 2 officially introduces the Space Capture feature, which allows users to record a space with the iPhone 15 Pro series smartphone and place it in the Vision Pro. In addition, iPhone Pro also supports "Obiect Capture", which allows you to grab a 3D model of an object by taking a circle** around it as an asset for later use. The iPhone has become a creative tool that makes it easier for people to take spaces with them wherever they go**, and the opportunities for 3D content production have greatly increased.
Apple Space** uses MV-HEVC encoding to help break down the barriers to sharing and interconnection between different devices. According to V2XR, Apple's official Video Toolbox SDK already provides support for MV-HEVC encoding. MV-HEVC encoding saves the information of the left and right eyes in each picture, and stores the difference information between the main view and the auxiliary view to improve storage efficiency. Compared to traditional 3D** using SBS (Side by Side) encoding mode, MV-HEVC is more efficient and can reduce bitrate by 20%. With universal MV-HEVC encoding, future iPhone shots will be viewable in the Quest browser, and Quest stereoscopic images will be shared with Apple Vision Pro users. This will drive the Phone to become the 3D** creation tool with the largest market share and the lowest barrier to use.
AUSD is an open-source, non-profit organization established by Apple Pixar, Adobe, Autodesk, NVIDIA, etc. At present, there is no unified format for 3D content, which can be applied on a large scale in various fields, and the fields of animation, film and television, games, industrial modeling, browsers, and mobile applications have generated demand for 3D models in different periods, and the inconsistent format has resulted in low content production efficiency and low output. We believe that AOUSD may be able to build a bridge to unify the format of 3D models in industries such as film and television, animation, and construction, and reduce the barriers to 3D content production.
Other giants are also actively promoting the XR content hardware ecosystem. According to "Wall Street**", Tencent has reached a preliminary agreement with Meta to exclusively bring Meta's "a new, ** lower VR headset version" in Chinese mainland at the end of 2024. Tencent has unique strengths in content and service operations in China, such as software subscriptions and game consumption, and Meta has advantages in hardware, and we believe that the cooperation between Tencent and Meta is expected to complement each other and promote the prosperity of the XR ecosystem. In summary, we believe that there are three important trends in the development of the XR industry in the near future.
Trend 1: Space** becomes XR rigid application. At present, XR applications are mainly games, and there is less ** content. Space** is an emerging form of content, which can provide a shocking effect far beyond the current 2D**, and has a better user experience. We believe that the improvement of user experience in space can be compared with the effect of user experience improvement in the era when it appears. Similar to the drive of shooting to smartphones, space is expected to become a rigid application for XR devices.
Trend 2: UGC has become an important growth driver for 3D content in the XR industry. At present, XR content production is mainly produced by professionals, the number of outputs is small, and the sense of user participation is low, and the UGC of space moments can not only enrich the diversity of ** content, but also stimulate users' creativity and participation. Referring to the development experience of YouTube, Facebook and other enterprises, when users can produce and share ** content by themselves, it can effectively drive user stickiness through social ties, thereby promoting the maturity of the XR industry. We believe that the threshold for content production from PGC to UGC 3D has been greatly reduced, which can effectively drive the amount of 3D content and user stickiness of XR devices.
Trend 3: Hardware ecological interconnection is expected to drive XR sales breakthroughs and promote the popularization of 3D sensing functions in smartphones. Hardware interconnection is embodied in: 1. Smartphones support shooting space**, support grabbing 3D models, and put space **on MR devices**:2. Different brands of XR devices support**space recorded with each other**. At present, the sales volume of smartphones is much larger than that of XR devices, and under the trend of hardware ecological interconnection, we believe that XR devices may be expected to achieve sales breakthroughs with the help of smartphone ownership. At the same time, we believe that this trend will also lead to the popularization of 3D perception in smartphones.
We believe there are three trends that are expected to accelerate the growth of 3D content, which in turn will drive sales of XR hardware.
In a simplified way, the 3D content industry chain includes three links: 3D content production, 3D transmission, and 3D content. As the giants push 3D content from the PGC ecosystem to the UGC ecosystem, the industry logic has changed greatly. PGC mode: The professional team uses special equipment to produce 3D content, the number of teams is small, the content transmission method is mainly distributed in the app store, without hardware intervention, and the downstream is based on high-performance XR all-in-one machines**, and the content form focuses on games. Under the three major trends mentioned in the previous section, in the future, smartphones may be used as hardware for 3D content production, and cable transmission methods may be added, and in the 3D content** link, affordable XR devices may also benefit from the space** and usher in rapid sales growth. We believe that under the trend of 3D content breakthrough and hardware interconnection, smartphone TOF modules, ** bridge chips, and affordable MR will usher in incremental opportunities.
Depth sensors can help Apple smartphones produce 3D content, with the TOF module at its core
At present, there are two types of 3D content for iPhone: grabbing 3D models of objects and recording space**. When the iPhone grabs the 3D model of the object (i.e., the Obiect Capture function), it needs to restore the depth information of the object with the help of the depth sensor to capture the XYZ three-axis information of the object, and quickly scan and model it, so as to maintain the 3D model in the phone: Spatial** recording is camera-based (i.e., the iphonc 15 Pro uses the main camera and the ultra-wide camera to record separately), but the distance between the two cameras on the iPhone 15 Pro is only about 2cm, which is lower than the average interpupillary distance (about 6cm), there is not enough distance between the two camera angles for stereo capture, and the depth sensor may assist in spatial** shooting. Therefore, depth sensors have a great impact on the production of 3D content on smartphones.
The existing 3D perception solutions for smartphones are mainly structured light and Tof. Due to the short working distance, the scattered light spots emitted by the structured light scheme are easy to be interfered with by outdoor strong light, which is more suitable for close-range application scenarios such as face recognitionCompared with structured light solutions, the TOF solution has better shooting distance and accuracy, making it an important helper for 3D content production. The rear LiDAR of the Apple iPhone and iPad measures the distance by measuring the time it takes for light to touch and reflect an object, using the direct time-of-flight (DTOF) method, to measure the environment up to 5 meters. The TOF module is an important layout for Apple to build an XR ecosystem.
*Bridge chips are currently used in split XR products and are in less demand. The principle of the bridge chip used in XR equipment is as follows: When the source is a mobile phone PC, by connecting to the AR VR display device, the signal format from the source is converted into different information from the left and right screens and sent to the display screen, and the signal needs to be bridged and controlled in the process. Typical products such as Rokid glasses, because their signal source comes from the host station, cables and ** bridge chips are required to transmit signals and format conversion, which can be smoothly *
The large volume of 3D content and the high stability requirements on XR equipment have led to the need for bridge chips in XR all-in-one machines. According to the 9to5mac, Apple's definition of space is shooting at 30 per second at 1080p, which requires about 130MB of storage per minute, and gigabytes of storage for just 10 minutes. Therefore, when smartphones have become an important means of 3D content production, how to better integrate mobile phone content on XR devices** has become an important bottleneck affecting the user experience. Due to XR's high requirements for the stability of content, if wireless transmission is used, there may be delays, frame drops, etc., which will cause users to dizzy, which will affect the consumer experience. Therefore, cable transmission may become an important way, and the possibility of supporting bridge chips has greatly increased.
* The bridge chip market is expected to open up greatly. From the perspective of XR brands, the added cost of adding a bridge chip is limited, but the benefits obtained are large XR devices will be able to directly capture 3D content captured by smartphones from convenience considerations, which can improve the user experience and expand the application scenarios of XR devices. We believe that brand owners will choose to carry the best bridge chip under the trade-off of pros and cons, and drive the best bridge chip into an incremental market.
Existing brand XR products have good performance and full functions, but they are expensive, and consumers have a high purchase threshold, which prevents consumers from entering the market. Among the mainstream products that have been released, the Apple Vision Pro is priced at $3,499, the Meta Quest3 is priced at $499, the Quest2 is currently priced at $249, the Pico4 is priced at 2,499 yuan, the Rokid Max is priced at 2,699 yuan, and the Xreal Air 2 is priced at 2,599 yuan. Affordable XR products have appeared on the market. ODM manufacturer E-realm Virtual released a thousand-yuan entry-level product A162 **only 1 25 of Vision Pro, which supports 3D video recording and space*** and other functions. It uses a Qualcomm XR1 chip and a nearly 4K 120-inch giant screen display. FOV up to 98°, 72Hz refresh rate, 55-inch fast-LCD screen with Fresnel optics. The A162 product introduces two cameras in front of the headset to support full-color VST capability. A162 Abandon the game scene and focus on watching movies and working. At present, Yijing has contacted the brand, and the ex-factory price of the product may be 149 169 US dollars.
Manufacturers of split AR glasses are also embracing space**. Due to the USB-C interface of the iPhone 15 Pro, the current split AR glasses can be directly connected to the mobile phone through a data cable, compared with the cumbersome operations such as VR products that need to be transferred through a PC, the AR glasses **space** is easier and the 3D experience is still more profound. China's leading AR glasses manufacturers are actively promoting the progress of 3D space** on the AR side, such as RokidIn early December 2023, we will promote activities such as the "3D Space Challenge" to develop the habit of users using their mobile phones to shoot 3D content and use AR glasses**.
The leading 3D visual perception solution provider in China is expected to benefit from the wave of 3D shooting of smartphones
The industry's leading 3D visual perception solution provider, the company has advantages in technology and customers. The company is the first enterprise in China to carry out systematic research and development of 3D visual perception technology, and has built two major technical sectors of 3D perception technology: Horizontally, the company has laid out a variety of 3D technology solutions, covering six technical directions, such as structured light, ITOF, binoccular, DTOF, Lidar, and industrial 3D measurementVertically, the company has independently developed a full-link technical system from deep warning chips, photosensitive chips, optical systems, SDKs, industry application algorithms, etc. The company has applied for more than 1,300 3D perception patents, expanding 3D visual perception to biometrics, consumer electronics, AIoT, industrial 3D measurement and other scenarios, serving more than 1,000 customers and many developers around the world.
Since its establishment, the technology has been rapidly selected, and the application scenarios have been continuously opened. Founded in 2013, the company focused on the field of artificial intelligence 3D sensing at the beginning of its establishment In order to grasp the opportunity of the era of the transition from 2D vision to 3D vision, the company has built a system of "full-stack technology research and development + full-technology path layout", and since 2015, it has independently developed and iterated a series of in-depth warning digital chips, developed a variety of photosensitive analog chips such as IOF and DTOF, and continuously developed 3D face payment, smartphone 3D camera, somatosensory 3D camera and other application directions, becoming a global possession One of the few companies with the ability of the 3D vision sensor industry chain.
3D visual perception has both software and hardware technologies, and perception chips and climbing algorithms are the core barriers.
The 3D visual perception industry chain is divided into upstream, midstream and downstream, the upstream is mainly laser, diffractive optical components, photosensitive chips, sensor modules, etc., and the midstream is dominated by 3D visual perception schemes, with deep warning algorithms, calibration alignment compensation and calibration algorithms and technical scheme design: The downstream is the algorithm for application scenario design, such as VSLAM, image segmentation, pose recognition and other specific schemes. The deep alarm chip solidifies the deep alarm algorithm, which can perform real-time depth calculation to output 3D data after receiving the spatial encoding information of the photosensitive chip, which belongs to the special ASIC chip: Compared with the traditional general-purpose processor invocation algorithm, the deep alarm chip has faster computing efficiency and more accurate computing power, which is the core of 3D vision productization.
The company's industrial chain layout is extensive, covering upstream photosensitive chips, sensor modules, midstream 3D perception algorithms, downstream application algorithms and solutions, and the main capabilities of the industrial chain have been built.
The company has laid out six main types of 3D perception technology solutions. 3D vision and 2D vision are greatly upgraded, and different scenarios require different 3D perception schemes. 2D images can only provide texture information, but cannot achieve the spatial topography, geometric size, posture and other information required for accurate recognition and tracking. 3D visual perception not only has texture information, but also increases depth information, that is, it can record spatial geometric size information, and the application scenario and value are greatly improved, and there are representative commercial applications in the fields of biometrics, robots, AIoT, 3D scanning, and industrial 3D measurement. The application scenarios of the company's products are as follows:
Different scenarios have different requirements for the measurement range, measurement accuracy, size, and power consumption of 3D vision, and different technical paths are derived to meet the needs of specific scenarios. At present, the main types of technologies are six categories, and the company has a layout in each technology path.
Building the advantages of full-stack technology R&D and mass production, the company's strength ranks in the first echelon of global 3D visual perception technology.
In terms of technology, 3D vision requires multi-level software and hardware R&D capabilities from photosensitive chips to background algorithms to solution integration, and there are various technical paths according to different program requirements. The company has built a 3D vision technology perception system of "full-stack technology R&D capability + full-field technology route layout", which has unique advantages in system design, chip design, algorithm R&D, optical system, software development, etc., and has R&D and innovation capabilities in structured light, ITOF, binocular 3D, DTOF, area array LiDAR and other technical paths. In terms of mass production, 3D sensors are precisely constructed and difficult to mass-produce. Relying on the early customer cooperation experience, the company has accumulated advantages in core equipment and technologies such as production process, testing tools and processes, calibration and alignment technology, self-calibration and compensation, and has built its own factory and put it into production, becoming one of the very few companies in the world that can mass-produce million-level area array 3D vision sensors except for Apple, Microsoft, Sony, Intel, Huawei, Samsung and other giants.
Revenue increased year-on-year, and losses narrowed. In the third quarter of 2023, the company achieved revenue growth, mainly due to the increase in sales revenue of the company's 3D vision sensor products. The year-on-year decrease in net profit attributable to the parent company was mainly due to the increase in the company's operating income and the decrease in R&D expenses. Because the company's 3D visual perception related products are still in the early stage of market development and have not yet been applied on a large scale, the company's revenue scale is relatively small, in addition, the company has maintained a high R&D investment, and the core backbone R&D, management and marketing personnel have been incentivized, which has brought a lot of expenses, so the company is still in a state of loss.
The company's gross profit margin remains at a high level, and the net profit margin level is expected to improve. The company's gross margin is relatively high, with a gross margin of 44 in the third quarter of 202306%。The company's net profit margin is still negative because: The company's R&D investment is high. In order to consolidate and expand the company's current market position and actively respond to the new challenges brought by future technologies and markets, in terms of R&D, technology R&D reserves have been carried out in accordance with the "full-stack technology R&D capability + full-field technology route layout", and the company's R&D personnel have invested relatively much since 2019, and the amount of R&D expenses and the proportion of revenue are relatively highThe company actively carries out equity incentives, and the cost is high. In order to further motivate the core backbone R&D, management and marketing personnel, the company carries out equity incentives through the employee stock ownership platform, recognizes a large amount of share-based payment expenses and includes them in the current profit and loss. The company is currently in a stage of steady growth, and the company's profitability is expected to improve as the advantages of scale and product technology differentiation are further highlighted.
Robots build the company's 3D vision long-term development space. Embodied intelligent robots will iterate in the direction of "brain first, perception breakthrough, and body improvement", and 3D, miniaturized, low-cost, high-performance and highly integrated perception systems have become rigid needs, and the development of embodied intelligent robot industry will drive 3D visual perception into an important stage before large-scale industrialization. The company has been providing 3D vision sensors for robots since 2015, and in recent years, it has launched a rich and comprehensive robot perception solution, with rich customers, and has reached cooperation with UBTECH, Tracking Technology, Stander Gaussian Robot, etc. In July 2023, the company released a fixed increase plan, planning to build a robot vision industry technology middle platform, and deepen the development of 3D vision perception technology in the field of robot applications.
Smartphones are expected to become an important increment in 3D visual perception, and the company has rich experience in mass production and is expected to fully benefit. In 2018, the company provided front-facing structured light technology for OPPO's flagship mobile phone, FindX, making it the second smartphone in the world to mass-produce one million units equipped with 3D vision sensors after the iPhone X: In terms of rear vision sensors, in May 2020, the company provided a system solution for Meizu's flagship mobile phone 17 Pro, bringing optimized 3D detection performance through its self-developed TOF depth engine and high-performance filtering algorithms, and in March 2021, the company provided a rear solution for Meizu 18 Pro ITOF's one-stop mass production solution improves the 3D perception performance of products through the technology empowerment of software and hardware integration. We believe that after the breakthrough of 3D content, XR+ smartphones are expected to form an ecological linkage trend, which will promote the rapid growth of demand for rear 3D vision sensors in smartphones. At present, only Apple's smartphones are equipped with rear 3D vision sensors on a large scale, and the market space is large in the future. With the company's accumulated technological advantages and mass production experience, it is expected to fully benefit from this trend.
Report total: 34 pages.
Featured Report**: The School of Artificial Intelligence