Abstract: A multimedia provider multicasts a main multimedia stream comprising a mixture of intra and inter frames and multicasts an associated auxiliary stream comprising intra frames. At least a portion of the intra frames in the auxiliary stream are generated based on same multimedia data as at least a portion of the inter frames of the main stream. The provider also transmits a stream information channel comprising information descriptive of time instances of intra frames in the main and auxiliary streams. This information allows a set top box to selectively connect, in connection with session set-up and channel switch, to the main or auxiliary stream to receive an intra frame with as a low delay as possible.

Abstract: The invention teaches a method of positioning a first image having a first image format into a second image comprising multiple image blocks and having a second image format. A straight boundary of at least one image block in the second image is identified. The first image is positioned into the second image by aligning an edge of the first image with the identified straight boundary. Alternatively, a block row or column of the first image can be aligned with the straight boundary but then in such a way that an edge of the first image parallel to the block row or column is aligned with a boundary of a row or column of image blocks in the second image. This image positioning reduces any bleeding artifacts and the number of bits required for representing the image during encoding.

Abstract: Today's computer programs that convert raster images into vector-based images do not optimize/compress the vector representation of the vector-based images. Instead, they simply keep all of the complex edges for the vector objects within the vector-based images. The present invention described herein functions to create a compressed vector-based image by simplifying the shapes of common complex edges which are shared by adjacent vector objects. The compression (lossless compression) of the vector objects is done without affecting the perceived quality of the vector-based image.

Abstract: A multimedia provider multicasts a main multimedia stream comprising a mixture of intra and inter frames and multicasts an associated auxiliary stream comprising intra frames. At least a portion of the intra frames in the auxiliary stream are generated based on same multimedia data as at least a portion of the inter frames of the main stream. The provider also transmits a stream information channel comprising information descriptive of time instances of intra frames in the main and auxiliary streams. This information allows a set top box to selectively connect, in connection with session set-up and channel switch, to the main or auxiliary stream to receive an intra frame with as a low delay as possible.

Abstract: Today's computer programs that convert raster images into vector-based images do not optimize/compress the vector representation of the vector-based images. Instead, they simply keep all of the complex edges for the vector objects within the vector-based images. The present invention described herein functions to create a compressed vector-based image by simplifying the shapes of common complex edges which are shared by adjacent vector objects. The compression (lossless compression) of the vector objects is done without affecting the perceived quality of the vector-based image.

Abstract: The invention teaches a method of positioning a first image having a first image format into a second image comprising multiple image blocks and having a second image format. A straight boundary of at least one image block in the second image is identified. The first image is positioned into the second image by aligning an edge of the first image with the identified straight boundary. Alternatively, a block row or column of the first image can be aligned with the straight boundary but then in such a way that an edge of the first image parallel to the block row or column is aligned with a boundary of a row or column of image blocks in the second image. This image positioning reduces any bleeding artifacts and the number of bits required for representing the image during encoding.