Are there any other alternatives to eliminate transaction malleability?

BIP62 was an attempt by Core developer Pieter Wuille. A number of other ideas have been floated such as new SIG_HASH types or Christian Decker’s “Normalized TXID” proposal, but these leave edge cases and amount to playing whack-a-mole with malleability issues. Others have suggested a complete overhaul of the transaction format which would require a hard fork and cause significant compatibility issues with existing software, pending time locked transactions,and existing outputs spendability .

Signature segregation is the cleanest way to deal with the problem but was never seriously considered because it was originally thought that this could only be implemented through a hard fork.