Transcript abundance quantification with IsoSeq

With IsoSeq data, the number of full-length reads matching a given transcript gives an approximate indication of the transcript's relative abundance. But the biases introduced by PCR, normalization (if done), size selection and loading add a lot of noise to the measurement.

Is there an accurate way to measure transcript abundance using IsoSeq data alone?

Alternatively, assuming you believe that Illumina RNASeq data suffers from no such biases, is there a way to combine IsoSeq and Illumina data to get at abundance?