diff --git a/Project.toml b/Project.toml
index 6625e5b..23cee31 100644
--- a/Project.toml
+++ b/Project.toml
@@ -4,8 +4,6 @@ version = "0.1.0"
 
 [deps]
 AbstractFFTs = "621f4979-c628-5d54-868e-fcf4e3e8185c"
-DSP = "717857b8-e6f2-59f4-9121-6e50c889abd2"
-DoubleFloats = "497a8b3b-efae-58df-a0af-a86822472b78"
 FFTW = "7a1cc6ca-52ef-59f5-83cd-3a7055c09341"
 LinearAlgebra = "37e2e46d-f89d-539d-b4ee-838fcccc9c8e"
 Reexport = "189a3867-3050-52da-a836-e630ba90ab69"
@@ -13,8 +11,6 @@ Test = "8dfed614-e22c-5e08-85e1-65c5234f0b40"
 
 [compat]
 AbstractFFTs = "1.0"
-DSP = "0.6, 0.7"
-DoubleFloats = "1.2"
 FFTW = "1"
 Reexport = "0.2, 1.0"
 julia = "1.7"
diff --git a/src/GenericFFT.jl b/src/GenericFFT.jl
index d7a06e1..62f90fa 100644
--- a/src/GenericFFT.jl
+++ b/src/GenericFFT.jl
@@ -2,8 +2,6 @@ module GenericFFT
 
 using Reexport
 
-import DSP
-
 @reexport using AbstractFFTs
 @reexport using FFTW
 
@@ -16,8 +14,6 @@ import AbstractFFTs: Plan, ScaledPlan,
                      fftshift, ifftshift, rfft_output_size, brfft_output_size,
                      plan_inv, normalization
 
-import DSP: conv
-
 import FFTW: dct, dct!, idct, idct!, plan_dct!, plan_idct!,
              plan_dct, plan_idct, fftwNumber
 
diff --git a/src/fft.jl b/src/fft.jl
index 8d0a856..678a666 100644
--- a/src/fft.jl
+++ b/src/fft.jl
@@ -81,11 +81,6 @@ function _conv!(u::StridedVector{T}, v::StridedVector{T}) where T<:AbstractFloat
     y = T<:Real ? real(y[1:n]) : y[1:n]
 end
 
-conv(u::AbstractArray{T, N}, v::AbstractArray{T, N}) where {T<:AbstractFloat, N} = _conv!(deepcopy(u), deepcopy(v))
-conv(u::AbstractArray{T, N}, v::AbstractArray{Complex{T}, N}) where {T<:AbstractFloat, N} = _conv!(complex(deepcopy(u)), deepcopy(v))
-conv(u::AbstractArray{Complex{T}, N}, v::AbstractArray{T, N}) where {T<:AbstractFloat, N} = _conv!(deepcopy(u), complex(deepcopy(v)))
-conv(u::AbstractArray{Complex{T}, N}, v::AbstractArray{Complex{T}, N}) where {T<:AbstractFloat, N} = _conv!(deepcopy(u), deepcopy(v))
-
 # This is a Cooley-Tukey FFT algorithm inspired by many widely available algorithms including:
 # c_radix2.c in the GNU Scientific Library and four1 in the Numerical Recipes in C.
 # However, the trigonometric recurrence is improved for greater efficiency.
diff --git a/test/Project.toml b/test/Project.toml
new file mode 100644
index 0000000..a03eaff
--- /dev/null
+++ b/test/Project.toml
@@ -0,0 +1,4 @@
+[deps]
+AbstractFFTs = "621f4979-c628-5d54-868e-fcf4e3e8185c"
+DoubleFloats = "497a8b3b-efae-58df-a0af-a86822472b78"
+Test = "8dfed614-e22c-5e08-85e1-65c5234f0b40"