Vidgyor works with TV broadcasters/OTT platforms using their cutting edge AI/ML Technology on different use cases for Broadcast TV & Streaming TV.
Our propreitary AI/ML based V-Spot technology can auto-detect TV ad breaks by analysing linear TV IP feeds using multiple heuristics & content fingerprinting techniques in near real time with high level of accuracy. V-Spot is used in multiple use cases like dynamic ad-insertion, broadcast monitoring etc.
V-Scan can analyze image, video & live stream content with highly scalable computer vision and machine learning technology. With V-Scan, you can identify objects, logos, people, face detection, text in images, videos & live stream for multiple use cases in broadcast media & other industry verticals.
Auto TV Ad-break detection using Artificial Intelligence & Machine Learning
Our propreitary AI/ML based V-Spot technology can auto-detect TV ad breaks by analysing linear TV IP feeds using multiple heuristics & content fingerprinting techniques in near real time with high level of accuracy. V-Spot is used in multiple use cases like dynamic ad-insertion, broadcast monitoring etc.
V-Spot auto-detects TV Ad-breaks without need of any Cue-tones/Ad-markers in the stream
V-Spot runs on cloud, can scale to 100's of TV channels. It works on any channel, any language, any genre
| TV Ad-Break Detection | Using AI/ML without Cue tones/Ad-markers |
| Cue-tones/Ad-markers | Not Required |
| Latency | Near real time |
| Deployment | Software which runs on cloud |
| Input | IP stream of Linear TV feed |
| Expense Model | Opex |
| 1. TV Broadcaster pushes IP stream of linear TV channel as input (with TV ads & without cuetones/ad-markers) to Vidgyor cloud. |
| 2. Vidgyor uses V-Spot to detect TV ad-breaks by analysing IP stream using AI/ML on cloud. |
| 3. V-Spot detects TV ad-breaks in near real time with high accuracy. |
| 4. Vidgyor auto-replaces original TV ads with targeted video ads on internet streaming devices using their V-Flows/V-Fusion/V-Alchemy DAI products. |
V-Scan can analyze image, video & live stream content with highly scalable computer vision and machine learning technology. With V-Scan, you can identify objects, logos, people, face detection, text in images, videos & live stream. With V-Scan, you can identify the objects/logos in live/video that are specific to your business needs.
For example; we can build a model to detect a logo and auto-replace the logo with different logo in near real time. Vidgyor offers V-Scan as managed solution, you simply need to supply training set of images of objects or logo you want to identify, and Vidgyor shall shall care of the the heavy lifting of model development for you & training the model for your specific usecase in broadcast media & other industry verticals.
With V-Scan, you can detect and classify multiple objects including the location of each object along with confidence score. For example you can identify objects (such as handbag, car, building), and scenes (such as parking lot, beach, city). When analyzing video, you can also identify specific activities such as "delivering a package" or "playing soccer" etc.
| Objection Detection | Multiple objects |
| Input Type | Image, Video, Live stream |
| Accuracy | Confidence score for each Object |
| Location | Location/Co-ordinates of Objects |
| Motion Analysis | Activities Identification |
V-Scan can auto-detect multiple instances of logo/object using computer vision & machine learning techniques from the live video and dynamically mask or replace with different logo on live video stream in near real time.
TV broadcasters use V-Scan to auto-replace TV sales sponsorships logos on laptops of news anchors, studio backgrounds and sponsor logos on screen with different set of sponsor logos for digital platforms. V-scan's logo masking is also being used for compliance related requirements while sending live feeds to other countries.
| No. of Objections | Multiple logo detection on single screen |
| Input Type | Image, Video, Live stream |
| Accuracy | Frame accurate replacement |
| Location | Location/Co-ordinates of Objects |
| Custom Logo Detection | Yes |
With V-Scan, you can detect text in photos, text appears very differently than neat words on a printed page. V-Scan can read skewed and distorted text to capture information like store names, street signs, and text on product packaging.
With V-Scan, you can easily detect when faces appear in images and videos and get attributes such as gender, age range, eyes open, glasses, facial hair for each. In video, you can also measure how these face attributes change over time, such as constructing a timeline of the emotions expressed by an actor.
V-Scan can also identify well known people (celebrities) in your live video, video file and image libraries to catalog footage and photos. It can also be trained to detect custom people.
With V-Scan, you can automatically transcribe video, extract audio transcript from live video/vod file for closed captioning,subtitling. Subtitles,Transcript can also be changed from one language to another.
Vidgyor's AI/ML technology trusted by 60+ TV Channels
302, 1st Floor, 3rd Cross
9H Main Road, HRBR Layout 1st Block, Kalyan Nagar, Bengaluru, Karnataka 560043. India
contact@vidgyor.com
+91 9886926820
+91 9449008000