With the development of information technology, computer-assisted language learning plays an increasingly important role in English teaching in colleges and universities. Traditional English teaching methods have problems such as insufficient interactivity and low degree of personalization, while computer-assisted teaching can provide a more flexible and diverse learning environment. In this study, a speech enhancement algorithm based on improved CTF-GSC and posterior Wiener filtering is proposed for the speech enhancement problem in computerassisted language learning and applied to the practice of English teaching in colleges and universities. Methodologically, the basic principle of Wiener filtering is first analyzed, and it is proposed to optimize the Wiener filtering enhancement effect by combining the wavelet thresholding multi-window spectral estimation algorithm with the VAD algorithm; Second, the VSS-NLMS algorithm is introduced to improve the CTF-GSC algorithm to further enhance the speech enhancement effect; Finally, two teaching modes, interactive and collaborative, were designed to realize a new model of computer-assisted English teaching in colleges and universities. The experimental results show that under Gaussian white noise environment, the speech enhancement algorithm proposed in this study improves the signal-to-noise ratio to 32.0173 dB, which is 20.5848 dB and 15.9994 dB higher than the spectral subtraction method and the traditional Wiener filtering algorithm, respectively; the improved algorithm obtains an average score of 4.02 in the subjective scoring test, which is higher than the other comparative algorithms; and in the actual teaching application, the experimental class students’ English scores on the posttest mean improved by 3.2 points over the pretest, while the control class only improved by 1.44 points. The study shows that the model of integrating computer-assisted language learning and traditional pedagogy can effectively improve the quality of speech and the effect of English teaching, which provides new ideas and methods for the reform of English teaching in colleges and universities.
