tencent cloud

Video and Media AI Solution
Developed by the top-notch Tencent R&D department, Video and Media AI Solution is an industry-specific solution backed by advantages in image processing and deep learning combined with leading technologies including face recognition, image recognition, ASR, OCR, TTS, and multimedia capabilities. It will fully meet your requirements for AI and scenario-based applications.
Contact Us
Strengths
Use Cases
Success Stories
Relevant Tencent Cloud Products

Strengths

Strengths

Use Cases

Success Stories

Relevant Tencent Cloud Products

Strengths
Leading Technologies

It is developed by Tencent based on rich data and tags accumulated from internal businesses in combination with in-depth business understanding.

Diverse AI Algorithm Capabilities

It integrates top-notch AI computing power and provides Intelligent video processing, video quality inspection, video filling, content production, and other applications. Its underlying AI algorithms set world records and fully optimize the production, management, and consumption of media assets. In addition, it offers a dozen of unique underlying capabilities for media businesses to output custom capabilities for specific scenarios, including lens jitter, low light, overexposure, and black/white/pure color/corrupted display recognition.

Industry-Leading Effect

From video preprocessing to algorithm capability cascading, it ensures the optimal efficiency through efficient orchestration and can quickly, accurately, and intuitively output more comprehensive video quality information in specific media scenarios, with both the quality inspection scope and effect leading the industry.

Wide Scenario Support

Optimized for the media industry and leveraging complete AI capabilities, it can comprehensively detect quality problems in video files and live streams in various business scenarios such as material search and content storage and production.

Use Cases

Intelligent Video Enhancement

Intelligent Video Quality Detection

Intelligent Video Content Prodution

Intelligent Video Impainting

Intelligent Face Fusion

Intelligent Video Enhancement

It provides an one-stop intelligent video processing service to repair various video quality defects caused by video storage, image, bitrate, frame count, or bandwidth during video production, transfer, and consumption. It also optimizes the video presentation quality, thus empowering the media industry to usher in the 4K era.

  • Video intelligent super resolution: it can intelligently remaster videos and enhance the video image quality to increase the resolution of videos.
  • Video intelligent transcoding: it can intelligently optimize the video quality and control the video bitrate.
  • Video intelligent remastering: overcoming the limitations of manual remastering technology, it can intelligently repair the damage to old videos, thus enabling the rerelease of classic videos.

Intelligent Video Quality Detection

With multidimensional AI quality inspection capabilities, it can thoroughly inspect videos and quickly detect and locate problems, allowing media personnel to easily check the quality of high numbers of video materials.

  • Video inspection: it can intelligently determine the video quality of a single frame or segment and output multiple video quality tags such as black bar, jitter, and smear.
  • Object detection: it can recognize barcodes or QR codes in videos.
  • Audio inspection: it can detect any exceptions in audios, such as crackling and low volume level.
  • Video scoring: it can score the quality and aesthetics of videos or audios based on the above quality characteristics.

Intelligent Video Content Prodution

It provides landscape/portrait mode conversion and AI-based ad placement capabilities and can adjust and optimize various business scenarios in a targeted manner, thus fully meeting your business needs.

  • Landscape/Portrait mode conversion: it can adaptively clip videos, making it widely and practically applicable in video editing and auxiliary editing.
  • AI-Based advertisement placement capability: it enables subpixel-level tracking and high-fidelity replacement of ads, so that media content can be produced and placed precisely.

Intelligent Video Impainting

Based on Tencent's leading deep learning and computer vision technologies, the Intelligent Video Impainting feature can efficiently remove specified parts in the video image to restore the occluded content. For example, it can remove static and dynamic logos and subtitles with a stable effect delivered. In addition, based on an extensive library of TV station logos, it can automatically recognize and quickly remove such logos to facilitate fast editing and production of video content in many application scenarios.

  • Intelligent logo removal: it can intelligently remove the logos in videos to restore the occluded content. It allows you to flexibly specify areas in videos for removal and can be used in combination with the intelligent TV station logo detection feature to locate logos for quicker removal.
  • Intelligent subtitles removal: it can remove subtitles in one or multiple specified areas in videos. Based on deep learning algorithms, it uses temporal information to easily fill specified subtitles with images, delivering a natural removal effect and quickly restoring the image areas occluded by subtitles.

Intelligent Face Fusion

It provides the portrait fusion technology, leveraging various AI technologies such as portrait keying, face fusion, and 3D face reconstruction, to extract and reprocess persons in images or videos, so as to increase the video production efficiency and expand the room for creative imagination.

Use Cases

It provides an one-stop intelligent video processing service to repair various video quality defects caused by video storage, image, bitrate, frame count, or bandwidth during video production, transfer, and consumption. It also optimizes the video presentation quality, thus empowering the media industry to usher in the 4K era.

  • Video intelligent super resolution: it can intelligently remaster videos and enhance the video image quality to increase the resolution of videos.
  • Video intelligent transcoding: it can intelligently optimize the video quality and control the video bitrate.
  • Video intelligent remastering: overcoming the limitations of manual remastering technology, it can intelligently repair the damage to old videos, thus enabling the rerelease of classic videos.

With multidimensional AI quality inspection capabilities, it can thoroughly inspect videos and quickly detect and locate problems, allowing media personnel to easily check the quality of high numbers of video materials.

  • Video inspection: it can intelligently determine the video quality of a single frame or segment and output multiple video quality tags such as black bar, jitter, and smear.
  • Object detection: it can recognize barcodes or QR codes in videos.
  • Audio inspection: it can detect any exceptions in audios, such as crackling and low volume level.
  • Video scoring: it can score the quality and aesthetics of videos or audios based on the above quality characteristics.

It provides landscape/portrait mode conversion and AI-based ad placement capabilities and can adjust and optimize various business scenarios in a targeted manner, thus fully meeting your business needs.

  • Landscape/Portrait mode conversion: it can adaptively clip videos, making it widely and practically applicable in video editing and auxiliary editing.
  • AI-Based advertisement placement capability: it enables subpixel-level tracking and high-fidelity replacement of ads, so that media content can be produced and placed precisely.

Based on Tencent's leading deep learning and computer vision technologies, the Intelligent Video Impainting feature can efficiently remove specified parts in the video image to restore the occluded content. For example, it can remove static and dynamic logos and subtitles with a stable effect delivered. In addition, based on an extensive library of TV station logos, it can automatically recognize and quickly remove such logos to facilitate fast editing and production of video content in many application scenarios.

  • Intelligent logo removal: it can intelligently remove the logos in videos to restore the occluded content. It allows you to flexibly specify areas in videos for removal and can be used in combination with the intelligent TV station logo detection feature to locate logos for quicker removal.
  • Intelligent subtitles removal: it can remove subtitles in one or multiple specified areas in videos. Based on deep learning algorithms, it uses temporal information to easily fill specified subtitles with images, delivering a natural removal effect and quickly restoring the image areas occluded by subtitles.

It provides the portrait fusion technology, leveraging various AI technologies such as portrait keying, face fusion, and 3D face reconstruction, to extract and reprocess persons in images or videos, so as to increase the video production efficiency and expand the room for creative imagination.

Success Stories
Yangshipin 5G new media platform
Tencent Cloud helped China Media Group launch Yangshipin, the first national 5G new media platform in China, and create a new flagship for audiovisual media by utilizing its advantageous resources.
Relevant Tencent Cloud Products
CVM
Cloud Virtual Machine (CVM) provides secure and reliable elastic computing services. It allows you to scale computing resources in real time to adapt to your changing business needs, and you only need to pay for what you use.
Learn more >
COS
Cloud Object Storage (COS) is a distributed storage service launched by Tencent Cloud. It has no directory hierarchy or data format restrictions, can accommodate massive amounts of data, and supports access over HTTP/HTTPS protocols.
Learn more >
Technical Consultation

To learn more about use cases and technical architectures, please contact and consult our sales and technical teams.