Transcoding Quality Prediction for Adap...

In recent years, video streaming applications have proliferated the demandfor Video Quality Assessment VQA). Reduced reference video quality assessment(RR-VQA) is a category of VQA where certain features (e.g., texture, edges) ofthe original video are provided for quality assessment. It is a popularresearch area for various applications such as social media, online games, andvideo streaming. This paper introduces a reduced reference Transcoding QualityPrediction Model (TQPM) to determine the visual quality score of the videopossibly transcoded in multiple stages. The quality is predicted using DiscreteCosine Transform (DCT)-energy-based features of the video (i.e., the video'sbrightness, spatial texture information, and temporal activity) and the targetbitrate representation of each transcoding stage. To do that, the problem isformulated, and a Long Short-Term Memory (LSTM)-based quality prediction modelis presented. Experimental results illustrate that, on average, TQPM yieldsPSNR, SSIM, and VMAF predictions with an R2 score of 0.83, 0.85, and 0.87,respectively, and Mean Absolute Error (MAE) of 1.31 dB, 1.19 dB, and 3.01,respectively, for single-stage transcoding. Furthermore, an R2 score of 0.84,0.86, and 0.91, respectively, and MAE of 1.32 dB, 1.33 dB, and 3.25,respectively, are observed for a two-stage transcoding scenario. Moreover, theaverage processing time of TQPM for 4s segments is 0.328s, making it apractical VQA method in online streaming applications.