Method of reducing context models for entropy coding of transform coefficient significant flag

A method of video decoding performed in a video decoder, the method comprising:

• receiving a coded video bitstream including a current picture and at least one syntax element that corresponds to transform coefficients of a transform block in the current picture;
• determining, for each context model region from a plurality of context model regions, an output of a monotonic non-decreasing function performed on a sum (x) of a group of partially reconstructed transform coefficients and a number of context models associated with a respective context model region;
• determining a context model index based on the output of the monotonic non-decreasing function of each context model region;
• and
• selecting, for the at least one syntax of a current transform coefficient, a context model from a plurality of context models based on the determined context model index.

A video decoder for video decoding, comprising:

• processing circuitry configured to:
• receive a coded video bitstream including a current picture and at least one syntax element that corresponds to transform coefficients of a transform block in the current picture,
• determine an offset value based on an output of a monotonic non-decreasing f(x) function performed on a sum (x) of a group of partially reconstructed transform coefficients,
• determine a context model index based on a sum of the determined offset value and a base value, and
• select, for the at least one syntax of a current transform coefficient, a context model from a plurality of context models based on the determined context model index.

A video decoder apparatus for video decoding, comprising:

• processing circuitry configured to:
• receive a coded video bitstream including a current picture and at least one syntax element that corresponds to transform coefficients of a transform block in the current picture,
• determine, for each context model region from a plurality of context model regions, an output of a monotonic non-decreasing function performed on a sum (x) of a group of partially reconstructed transform coefficients and a number of context models associated with a respective context model region,
• determine a context model index based on the output of the monotonic non-decreasing function of each context model region, and
• select, for the at least one syntax of a current transform coefficient, a context model from a plurality of context models based on the determined context model index.